{ "cells": [ { "cell_type": "markdown", "metadata": {}, "source": [ "# Introducing Scikit-Learn" ] }, { "cell_type": "markdown", "metadata": { "deletable": true, "editable": true }, "source": [ "There are several Python libraries that provide solid implementations of a range of machine learning algorithms.\n", "One of the best known is [Scikit-Learn](http://scikit-learn.org), a package that provides efficient versions of a large number of common algorithms.\n", "Scikit-Learn is characterized by a clean, uniform, and streamlined API, as well as by very useful and complete online documentation.\n", "A benefit of this uniformity is that once you understand the basic use and syntax of Scikit-Learn for one type of model, switching to a new model or algorithm is straightforward.\n", "\n", "This chapter provides an overview of the Scikit-Learn API. A solid understanding of these API elements will form the foundation for understanding the deeper practical discussion of machine learning algorithms and approaches in the following chapters.\n", "\n", "We will start by covering data representation in Scikit-Learn, then delve into the Estimator API, and finally go through a more interesting example of using these tools for exploring a set of images of handwritten digits." ] }, { "cell_type": "markdown", "metadata": { "deletable": true, "editable": true }, "source": [ "## Data Representation in Scikit-Learn" ] }, { "cell_type": "markdown", "metadata": { "deletable": true, "editable": true }, "source": [ "Machine learning is about creating models from data: for that reason, we'll start by discussing how data can be represented.\n", "The best way to think about data within Scikit-Learn is in terms of *tables*." ] }, { "cell_type": "markdown", "metadata": { "deletable": true, "editable": true }, "source": [ "A basic table is a two-dimensional grid of data, in which the rows represent individual elements of the dataset, and the columns represent quantities related to each of these elements.\n", "For example, consider the [Iris dataset](https://en.wikipedia.org/wiki/Iris_flower_data_set), famously analyzed by Ronald Fisher in 1936.\n", "We can download this dataset in the form of a Pandas `DataFrame` using the [Seaborn](http://seaborn.pydata.org/) library, and take a look at the first few items:" ] }, { "cell_type": "code", "execution_count": 1, "metadata": { "collapsed": false, "deletable": true, "editable": true, "jupyter": { "outputs_hidden": false } }, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
sepal_lengthsepal_widthpetal_lengthpetal_widthspecies
05.13.51.40.2setosa
14.93.01.40.2setosa
24.73.21.30.2setosa
34.63.11.50.2setosa
45.03.61.40.2setosa
\n", "
" ], "text/plain": [ " sepal_length sepal_width petal_length petal_width species\n", "0 5.1 3.5 1.4 0.2 setosa\n", "1 4.9 3.0 1.4 0.2 setosa\n", "2 4.7 3.2 1.3 0.2 setosa\n", "3 4.6 3.1 1.5 0.2 setosa\n", "4 5.0 3.6 1.4 0.2 setosa" ] }, "execution_count": 1, "metadata": {}, "output_type": "execute_result" } ], "source": [ "import seaborn as sns\n", "iris = sns.load_dataset('iris')\n", "iris.head()" ] }, { "cell_type": "markdown", "metadata": { "deletable": true, "editable": true }, "source": [ "Here each row of the data refers to a single observed flower, and the number of rows is the total number of flowers in the dataset.\n", "In general, we will refer to the rows of the matrix as *samples*, and the number of rows as `n_samples`.\n", "\n", "Likewise, each column of the data refers to a particular quantitative piece of information that describes each sample.\n", "In general, we will refer to the columns of the matrix as *features*, and the number of columns as `n_features`." ] }, { "cell_type": "markdown", "metadata": { "deletable": true, "editable": true }, "source": [ "### The Features Matrix\n", "\n", "The table layout makes clear that the information can be thought of as a two-dimensional numerical array or matrix, which we will call the *features matrix*.\n", "By convention, this matrix is often stored in a variable named `X`.\n", "The features matrix is assumed to be two-dimensional, with shape `[n_samples, n_features]`, and is most often contained in a NumPy array or a Pandas `DataFrame`, though some Scikit-Learn models also accept SciPy sparse matrices.\n", "\n", "The samples (i.e., rows) always refer to the individual objects described by the dataset.\n", "For example, a sample might represent a flower, a person, a document, an image, a sound file, a video, an astronomical object, or anything else you can describe with a set of quantitative measurements.\n", "\n", "The features (i.e., columns) always refer to the distinct observations that describe each sample in a quantitative manner.\n", "Features are often real-valued, but may be Boolean or discrete-valued in some cases." ] }, { "cell_type": "markdown", "metadata": { "deletable": true, "editable": true }, "source": [ "### The Target Array\n", "\n", "In addition to the feature matrix `X`, we also generally work with a *label* or *target* array, which by convention we will usually call `y`.\n", "The target array is usually one-dimensional, with length `n_samples`, and is generally contained in a NumPy array or Pandas `Series`.\n", "The target array may have continuous numerical values, or discrete classes/labels.\n", "While some Scikit-Learn estimators do handle multiple target values in the form of a two-dimensional, `[n_samples, n_targets]` target array, we will primarily be working with the common case of a one-dimensional target array.\n", "\n", "A common point of confusion is how the target array differs from the other feature columns. The distinguishing characteristic of the target array is that it is usually the quantity we want to *predict from the features*: in statistical terms, it is the dependent variable.\n", "For example, given the preceding data we may wish to construct a model that can predict the species of flower based on the other measurements; in this case, the `species` column would be considered the target array.\n", "\n", "With this target array in mind, we can use Seaborn (discussed in [Visualization With Seaborn](04.14-Visualization-With-Seaborn.ipynb)) to conveniently visualize the data (see the following figure):" ] }, { "cell_type": "code", "execution_count": 2, "metadata": { "collapsed": false, "deletable": true, "editable": true, "jupyter": { "outputs_hidden": false } }, "outputs": [ { "data": { "image/png": "iVBORw0KGgoAAAANSUhEUgAAAgMAAAGrCAYAAABdfJRJAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjQuMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/MnkTPAAAACXBIWXMAAAsTAAALEwEAmpwYAAEAAElEQVR4nOyddXgbV9aH3xFbkiVLZmY7DjM3aZukzMztlnELC6XddnfL3XK7X5mZmdKkkKZhBjsxM7MFFs73xzhyHDmJncSxncz7PHps3aEzutLMmXvP+R1BFEVkZGRkZGRkDl8Ug22AjIyMjIyMzOAiOwMyMjIyMjKHObIzICMjIyMjc5gjOwMyMjIyMjKHObIzICMjIyMjc5gjOwMyMjIyMjKHOQPuDAiCcKsgCFsFQdgiCML7giDoBvqYMjIyMjIyMn1nQJ0BQRDigT8Dk0VRHA0ogfMG8pgyMjIyMjIy/eNgTBOogBBBEFSAHqje3YrHHXecCMivgXv1mUO9LxodjWKjs3HI9wMc+n3Rn5fX7xVr7bViu6td7osBfDU7m8V6R/3BPq7MIKIayJ2LolglCMJjQDngBBaKorhwd+s3NjYOpDky/eBQ7YuWzha+Lf6Wlze/jEJQcP346zk2+VhMWtNgm7ZbDtW+6C8VHRW8ueVNvir+igRjAn+b8jemxkxFqVAeNBsO9b6we+wsLl/MM+ueodPXyZ9G/YnT0k8jXB8+2KbJDDADPU1gAU4FUoE4wCAIwkW7rHO1IAhrBEFY09DQMJDmyOyFw6Evllcv55HVj9Dc2Uyjs5H/LP8Pa+rWDLZZQRwOfdEfPD4PL296mQ/zP8TpdVLQWsD1i65ne8v2AT/24dQXG+o3cPfSu6lz1NHmauOpdU/xS+Uvg22WzEFgoKcJ5gMloig2iKLoAT4DZu68giiKL4miOFkUxcmRkZEDZogoivj98kjUnjhYfTFY+EU/nxR8EtT+fcn3g2DNnjnU+6K/1Dnq+Kroqx5tXtFLcVvxgB/7cOqL3yp/C2r7aPtHOD3OQbBG5mAy0M5AOTBdEAS9IAgCMA/IG+Bj9sotH27gqMd+xePzD8bhZYYACkFBcmhyUHtiaOIgWCPTH3QqHeEhwUPVRrVxEKw5dIkzxAW1JYYmolaqB8EamYPJQMcMrBQE4RNgHeAF1gMvDeQxe6Ohw8XivHoijBr+KGzkyOyog22CzG6wuW3kNudS1VFFtD6anPAc3D43uc25tLnaSDWnkmPNocpWRV6T5EdmW7NJD0vfp+OdlXUW35V8h8PrAMCkMXFMyjEH7HxkDiy19lrymvJoc7fxz2n/5On1T1PYWgjA6PDRKFDw4bYPiTPGYXPbsOgsjAwfOaRjQIYys+Nn8/rW12nubAZAq9RyyahLcHgc5DblUmOvIdWUSmlHKQ6Pg1RzKhOjJqJVaQfZcpn9ZUCdAQBRFO8F7h3o4+yJZUWNjIozkRSuZ6nsDAwZPD4P7297n2fWPxNou2f6PXxX8l2PefwX5r/A7b/fTpurDZCeBl899lVGho/s9zGVCiVXjrkSt88NAqgVapQcvAA0mb5Tb6/n9iW3s65+HQACAg8d8RA+vw+lQklJWwk3/nwjIiIqQcVtk2/jrj/u4vLRl3P1mKvlG9Q+oBAUXJhzIV6/F1EU0aq0KFHyZu6bvLTpJf4z8z/ct+I+8lvzA+s/OudRjk05dpAtl9lfBtwZGAqsL28hLdJISrieRXl1g22OTBflHeX834b/69HW1NnUwxGI1kfzc/nPAUcAwOax8XnB5/vkDHyS/wkfbv+wR5vH5yHTmtnvfckMLHkteQFHAEBE5Im1T/DhSR/yS/kvvLjpxcAyr+jlm6JvmBM/h1c2v8IxyceQbc0eDLOHNUurlvLs+mcByfkSEclLzqPGXgOA1+8NOAIgxeE8t/45xkaMJdYYOyg2yxwYDgs54i1V7aRGGEiy6smvsw22OTJd2D12vKK3R5vX3/O9RWeh1lEbtG1RaxF+sX/xH37R32vAWUl7Sb/2I3Nw6HB1BLU1OhtpcjYFhrF3ps5RR3hIOH7Rj91jPxgmHnJUdFQE/he7Uv/LO8qx6qwAgem1nalz1PXaLjO8OOSdAVEUya/vINESgjlEjc8v0mRzDbZZMkCCMYEUU0qPNoPagEroHrAqbC1kZuxMdmVe8jwUQv++vgpBwRkZZwS1H59yfL/2I3NwSAtLQyn0nMI5PvV4ituKSTYFB4IelXgUy2uWE2OIISE04WCZeUgxJ2FOUNuC5AVU2aoAiDHEICD0WH5cynEkGOXPe7hzyDsDzXY3fj+YQ9QIgkBcmI7SJvmpYShgDbHy2NzHmBk3E6WgZHzkeKbFTOO5ec+RakpFpVBxYuqJROujuXz05Zi1ZkLVoVw88mLqHfX9HhkAmBk3k9sm3UaoOhSz1sydU+9kcszkATg7mf0l25LNc/OeI8WUgkqh4rSM07h+3PUsLF3IwtKF3DvjXmIMMWiVWk7POB2dUkecIY5njnqGKL0cF7QvTIiawL9n/ptwXTh6lZ7zss+jor2CMzLPYHzkeL4q/Ir7Z91PvDEetULNKemncEHOBXJ8xiHAIR8zUNJoJz4sBCmzEaJNOsqbHUxKtg6yZTIgZQY8deRTtLhaMGlMGDVSqtjbJ7yN0+skXBfOZ4Wf8VXRV5yYeiIKQcHC0oVckHNBv0cGQHJA/jT6T5yYeiIIyDeNIYxSoWR2/GxGHT8Ku8dOjD4GlVLFzPiZ/Gf5f9jUuImrx16NUW0kTBtGijmFMG0YerV+sE0floiiiFqp5ozMM5gZO5NfK3/lza1vUmmrRFGsYFbcLK4YcwXZ1mwmxUzC6XESHxpPiCpksE2XOQAcFs5AjLm7UGK4QUN5kzy/NZQIUYcQou6+oOQ25fLR9o/Y3rKd0zJOY3LUZMaEj+G9be8BMD12OguSFuzXMaMMshMwHMhtyuXDbR+S35rP6Rmnc1TiUcyKm8XRiUfzc8XP3LfiPqbFTuPeGfcSZwzOkZfpG/kt+Xya/ykbGjZwYuqJLEhewMSoibyT9w4gOQrpYemkmdMwqA0Y1IZBtljmQHPIOwNlTQ4ijJrA+wijlvJm2RkYqpS0lXDlj1fS4ZGCx7Y0buGK0VfwwOwHKO8oxy/6STYly3nkhwHFrcVc8eMV2DxS0O+Wxi3UOeq4YfwNPDD7Acray+TvwwGgqqOKa3+6lganJLWc25RLQUsB/5j+D9449g0qbBXoVXqSTcnoVHIF+kOVQ94ZKG2yk2jpHjaMMGrZXNW2hy1kBpP8lvyAI7CDt3PfZnrsdJJNyXL60mFEfkt+wBHYwZtb32RazDRSzCmMihg1SJYdGtjcNjY3bqbaVh1wBHbwZdGXXDb6MtLD0onQRwyShTIHkz5PugqCkCUIwsuCICwUBOHnHa+BNO5AUNHsIMrUHdwSYdRS09Y5iBbJ7Ildo8cBVAoVv1f9zh2/30GD49AuFCPTTW/VCNUKNb9V/saDKx/soT0h0z9EUeSLwi+4+qerqXMEa68oFcp9ismRGb70Z2TgY+AF4GXANzDmHHgqW5xEGrudgXCjhrr2TkRRDAQVygwdRlhHEKOP6aEtcFbWWagVaqbGTmVd3TrGR40n2hBNWVsZZR1lGNVGMsIyeh0qbne1U9haiM1jIzk0mWRzcEqazNCioqOC0rZSLFoL0froHjers7PO5ruS76hz1HHJyEuYGD1xEC0dPrS52ihqLcLusZNsSkYhKHh/2/tcO+5akk3JpJpTKWnr1tu4ZOQl/UrPFEWR4rZiKjsqsegspIely3EFw4z+OANeURSfHzBLBoBOj4/2Tg8WQ3fMgE6tRK1U0Oro2S4zNEgITeDFBS/yc/nPbG3aygjrCKw6K+/kvRMQDEoOTebfs/7NjYtvDAwjn5R2En+d/NcexWyanE08tuYxvin+BpBkjJ+f/zzjo8Yf9POS6Ru5jblcs+gaWl2tqBVqbp9yO7WOWsrby5kVNwudSse7ee8C4PTKlfT6QqOjkUdWP8IPpT8AEKoO5emjn+bc7HN5at1T+EQfl4++HJWgoqKjgiMTj2RyzGTUir4XJ1pWvYw///xn3H43AJeNuoyrx15NqCZ0QM5J5sCz13EgQRCsgiBYga8FQbheEITYHW1d7UOWqlYnkaFaFLuMAEQYNdS2y1MFQ5W0sDQuHHkh0fpoPtj2AcVtxT2UA8s6yvi+5PtAGiLAN8XfBAoZ7SCvKS/gCIAkY/zf1f+lwx2sbCcz+HR6O3luw3O0uloB8Pg93L/yfswaM0VtRSypWsKiskUck3IMVp2VFHPKoNo7XMhtyg04AgAdng6eW/8ceU15ePwe/KKfVza/wnvb3uO8EedxXOpxRIRIcQJ2j53ClkIqOioQxd5LwDc4Grh32b0BRwDgja1vUNBSMLAnJnNA6cvIwFpAhIDs1N92WiYCaQfaqAPFrlMEO7AaNNS0OcmJlSOQhyohqhDOzzmfGEMMK2tXBi3Pb8knwZhArb17OmHXIKhd3wPkNufS4e6Qn1iGIDaPja1NW4Paax211NhqcHldjI0cy8SoiVw88mLijfGDYOXwozc577zmvKAAzHZ3e4+YnNK2Uh5a9RDLqpcRogrhlom3cGrGqUHD/+3u9l7jDnr7/ckMXfY6MiCKYqooimlATtf/gRfQ/0oxB5GqFicRvTgDYXoNtW2yJPFQJ9mUzMUjL2ZOfLBE6pToKRS09nzy2HWOs7c5zznxcwI66zIHCbcdnK17Xc2sNXNU4lFB7VadFYfXwaToSQgILEhesE9Fqg4rnK3glqZRepNunpMwhyZnU482AYH4UMnB8vg8vLblNZZVL5N253Xy0KqH2NoY7KxF6CLItgQXhZIliocX/QkXXdbHtgCCIGQLgrBhp1e7IAi39MvC/aCi2YHVGBwXEKZXU9smzzcOB5QKJS6fi2OSjwloos9NmMuYyDHE66U0Q41Cw98n3EyOJafHtjnWHP42+W9oFNJ3YKR1JDdNuEnOlT5Y+LxQ/Bu8eza8Mg9WvQy2+t2urlaouWTkJYyNGAtIWSTnZZ/HhvoNTIqeREZYBpePuZwwXdhBOoFhSHsNLP8fvHI0fHAelC5jlDWH2ybdFvgdjI4YzXXjruPikRcHbtgGtYEHZz9IRlgGAM2dzfxU9lPQ7ovaioLazDoz/5n5H5JCkwDQq/Q8MOsBMi1yJdDhxF6nCQRBiAHigRBBECbQPV1gAvao+ymK4nZgfNd+lEAV8Pl+2NsvKlocPTQGdmDVa6iW0wuHDT7RR72jnmvGXgMCbG/aTl1bOS/psqgaMQe9x0nS7y+jiJwKsWMD2xk1Ri7MuZA5CXNweB3EG+Mxa82DeCaHGTUb4J3Twd+VfLTwblDpYOLFu90kLSyNC0ZcwOyO2ehVegxqA+MjxxOmDWNUxCjZEdgTogjr34ZfHpDeNxVB5WqMV/zEJSMv4cjEI+n0dhJvjA9k3rxzwjvU2msxaU0khiYGdmXUGMmyZLGufh3R+mjsHjs2j40YfUyvhx4ZMZK3j3+bWkctoZrQHvuSGR70JWbgWOAyIAF4Yqf2DuCufhxrHlAkimJZP7bZLypbnExMsgS1Wwwa8mraD5YZMvvJ0YlH81nBZ7yw6QXun3U/RpWRvPZitKmTMSi1NNZvxj/uLNLq83o4AyCNLMiBZoNE5epuR2DSn8AQgaOjio1F37GsOZdYYxyjI0aztWkr1bZqpsVMQ6VQsalxE0a1kRZXC0+se4IEYwLPHf2c7AjsjY5aWPF/3e8nXASmeFj9Csrk2aQmzYDw1B6bhIeE98jA2YFBbeCvk//Kmro1FLcVY9KYSDYlMyp8FJsaNrGyZiV1jjqmx05nUtQkLCEWrCFWrCHyFNxwZa/OgCiKbwJvCoJwpiiKn+7Hsc4D3t+1URCEq4GrAZKSkvZj98FUtzp7SBHvwGrQUNcuxwzsykD2xf6QGpbKq8e+Sp29lodWPRwIMvu04FPOzT6XFmcLrzas59Vp/+JQUREYqn3RL3Zke4w4EVpKIPcLfp73F+5centglciQSE5JP4U3tr7BG1vf4MbxN/JN8Te0u9tJM6fx6jGvEh8aT6xh8JQnh01fKNWgDQVnC2TMB3sDrJdqC7DmNRh9Fpz0FOj6Fjxbba/mibXdz3/hunCyLFnc9uttgeDAD7d/yD+n/5Nzss850Gcjc5Dpj85AsiAIt+3S1gasFUVxw542FARBA5wC3LnrMlEUXwJeApg8eXLvuSv7gNvrp8XhxmroPZugvkOeJtiVgeqLfmNvgrot0kUtPAOicvCJPmrstQFHIDIkkrOzzsYn+pgYPRG1Uk2ep+2QcQaGTF/sD4lTwRAB2SdAezVNky7hqdyeUiUNzgY0ym6H/ZOCTzg2+ViMGiNalZZae+2gFyAaNn1hiIDjHobaTaCPAGczhFhg4wfS8i2fwIwbIH7vQk0tnS08te6pHm1NnU3kt+QzLXYayaZkvH4vSkHJx9s/ZkbsDBJN8tTAcKY/zsDkrtfXXe9PAjYB1wqC8LEoio/uYdvjgXWiKAbnnwwQNW1OrAYNSkWwymCoVkWnx4/T7SNEEyx5KjOI2Brg+7/B1q7QEoWK4su/4fqV/+bKMVcCUsDg5aMv58m1TwZymydHT0avk+MBhhSR2XDxV/D747D1U7yzbsDhCS4StnP+usPjYHLMZP675r80OhsBiDPG8fy850kLG7JZzEMHrRmWPgXeroedxGkw8VJY96b03ufe7aY74/V7cXqCg6wtWgvVtuqAfodGoeEvk/+C1+89ENbLDCL9cQYSgImiKNoABEG4F/gWmIOkRbAnZ+B8epkiGEgqW5xEhQaPCgAIghCQJU6JkCUzBwOPz0N5Rzmd3k4SjAmYd9zIazd3OwIAeiuLy38hzhhHZEgkf57wZ3x+H58WfBpwBEaFj2JyzGR8fh/19nq5PPFgY2uA1nLQmcBtg4h0mPt3olx2Ls04g+fy3uTIxCMZYR2BAgXR+mjMWjNnZJzBzLiZ1NhrOC3jNH4o+YFKWyXVtmoWly+WnYG90dkOP/2z2xHQWyFhMsSMAa0RqteDNR1cHdBcDAo1DYYwap1NmDQmkkxJAYl2vUrPKRmn8PqW1wO7VyvUePwe2txtXD32agQEqm3VfFn4JdNjplNrq6XB2YBJYyLRlCjXNhhm9McZiAJ2nmj3ANGiKDoFQdjtBLwgCAZgAXDNvpm4b1Q0O3rVGNhBuEFDdZtTdgYGgbbONt7Je4dXNr+CV/QyKnwUD8x+gPSwdHA09lw5NBaNKRZjZy03/nwjIKUWzoidQWFrIQuSF6BX6Xlp00v4RT/jIsZxx7Q7GB0xehDOTIaaTfDxZdBcBDkngzkRVr0Ifh9C3ETOSL6FzFkP8nbhZ7yw8QUEBE5KO4l/z/g39Y56/rX8X1TZqtApdVwy6hJ+r/ydvOY8tjRtGewzG/q4OqBxu/S/IRJm3SyNyjhbpOmCU/8HHgd8eT2U/MbmEx/mtqL3qbXXEqIK4a5pd3FC6glolBpsHhs5lhyuGXsNi8sXE64L5/SM0wGYFjON17a8htfvJT0snTMyz6CwrZDH1zxOtb0anVLHndPu5MTUE9Gqdn8Nlhla9Md1exdYKQjCvV2jAn8A73Xd7HN3t5EoinZRFMNFUTyoJcbK9+IMWA0aauX0wkFhS9MWXtj0Al5RGlrc2rSVVze/isfnkZ5cdpaPdraiQOCXil8CTb9V/oZOpcOqs5JlyeLLoi/xi34ANjZu5JP8T3B7+zYcKnMA6WyH7/4OShVMvQpS50jR7TsyCqrXEVn0K9saNrGmbg0AIiJfF39Nla2Kzws/p8pWJe3K18lLm15ifvJ8ABYkLxiUUxpWKFSQeaz0/4SL4NeHJUcApL+fXyONDhQspGXsWdxd8llAwdPpdXLPH/cEJISNGiM/lP7A+9veJ9mUjE/0cefSO1EpVLy37b3AtEBRaxErqlfwS8UvVNurAanv7l12b5AomMzQps/OgCiK9yE93bd2va4VRfE/XTf7CwfGvH2nrMlB5G6mCQAseo1cyniQKGoNFi5ZUrVE0qSPGQNnviYNcQKMu5A1tWuC1l9Tt4YzM86kxl4TtGxFzQoaOxuD2mUGGHsjJE2F2PGw/UdoCc4idngdLG5YF9S+rWUbdo89qN3lc3HhiAuZETNjICw+NHB1wKaPpYyBESdB+tGgUEpTNLuu1yJVJmy0JFHS0bN/RESqbdIN3ea2sap2Fe3udhaXL2ZN3RpEROrswWFfK2tW9pqeuMOxkxke9HdSZx1SKePPgXpBEIZsnk15s4No0+6V5iwGDZUtwcFMMgNPb4IkYyPGSvUCVBoYfQZc9TOc+CTYaphkHQFIsqqpJilPOiMsgzpHXa8yqCPDR2LVyvnOBx2dCVrKYdOH0F4pvd91FdRMCZP6M82cFpCMTjOn9VolL8eaw5yEOYTrg282Ml2Ur4TProT2Kij9XRIfihoJyl3SqpUaSXcAMNuaiNJHoVVqybZkByS6I/WRAJg0Jo5JPiboUBH6iKC2nPAcWjtbg9qjQuTYneFEn2MGBEG4CbgXqAN8SEqEIjB2T9sNFpUtex4ZCDdqWFXSfBAtktnB6IjRHJtyLD+W/ghI2vM3TrixWya4Pg8atsO3t0LMGOZOOB/d9CjW1a3DJ/q4IOcCPD4PT617imNSjmFazLRAMaNofTQX5VyETi1LDh90XB2w7Svpf9EPLaXSU2rRz1KbMQpbxpHMN0WRGDmSdXXr0Cq1XD76cuL0cWiztDy7/tlAaeLzR5xPo7ORo5OOHpzzGS5s+1b6qzHChnfB1Q6tZTD7Vvj9MWmaRqGEEx6TRm2s6URt+IBHzn6eVR3lbG3eyqToSUyMnkimOZPl1cv5sfRHPH4Pf5/yd74s/JLtLdtZkLwAr8/LCakn8F3Jd4DkNJyQegJ+0c/CsoWBvrt67NWyHPEwoz8BhDcD2aIoNu11zUGmo9OD0+MjLGT39bgjjFpqWuVpgsEgUh/JPdPv4cKcC3F6nSSbkrsr0DUXw1unwdizuwRUWikTnTy48kF8ojT3/GPpjzw29zEuH3M59624j/Oyz+PUjFNRK9RkhGWQYckYvJM7nFHppEA1e9cUzYb3pGHr01+UAgqjR/Oj4MDRUcJjax4LbPZD6Q/8dfJfUQgKHp3zKDaPDZPGhNPjZFHZIiZGTSRKLz9l7hZzV0EgZxMYIyVnoLkYtn4Gc/4OodHQUQeVa2DkqXDZt3hay1lSu4TX894K7GZp1VJM003csPgGPH4PAF8Xf81/5/yXdlc7v1X9xl1/3MWsuFncNfUu4oxxxBnjaO1sRaVQ8dZxb9HgbMCqs5IWlkaIKmQwPg2ZfaQ/zkAFksjQkKesyUGsOSSQJtMbEUYtNe1ORFEMXs/vl4bbjNEQNWKArT08MWlNTIia0KOtsLWQgvp1RJ3zEjXuVjoSchgRlslPxV8HHAGQ5ja/Lf6W87LPw6K10O5uZ339eibHTJYdgcFEGwrz7oWvbupuaymFilWw8X0aLniPP0p+odXd2mMzj99Dta2a9LB0Or2dhChC6PR08tclfwVgatxUsqxZB+88hhvZx4PPJU0DpM6Fr26UpgoaC6QpmxEnwrJnpXUnXQoJk6kSXazctIabJtyEy+dCo9DQ4Gzgj6o/Ao7ADj7a/hGXjbqMeYnzGGEdQZwhDlEUaXA2EBESwcToiSgVkl7LCOTr5XClP85AMfCrIAjfslOKoSiKT+x+k8GhtMlOtGnPKS0GjRJRhHanF7N+lxGEr/8MpUslD3vuHTDt6gG0VgZgS+MWLv/xcm6ecDMfbH2NdfVSkFlGWAYjrMEXGJfPxXMbnmNq7FS+L/kek8bE0QlHHmSrZQL4vLD2DUnlbt49UmZBaCyEJUL1Oph9K2JLOYIg9CpQ0+5u538b/kedo45rx16LX/RzeubpfF7wOT6/L/h4Mt0IClj3NnRUQ+QImP9vKbPA1S6Vj965XkHXZ68QFByZcCTPrn82sOjE1BNR64JHU91+N6Xtpfx3zX8DbWdknkFhSyH3N93P/+b9j1nxswbu/GQOCv1xBsq7Xpqu15CltNG+W8GhHQiCQLRJR0WLA7N+J+W6kt+haDGc9Ay42uD726XRgdQ5A2z1IYzfB02FYK+H0DiwpuFwtlLash2n10GCOY33895nQfICQjWhOL1Orh93PT7Rh8PrINGYyLfF3yLSrVQ3PW46T659koLWAi4YcQGvbnkVh8/J9oYttHkdROmjSDIlycInB4vmYlj8b+np1N4ojRJ0toLLCkoNfkMkztixnBWVQbOrmY0NGwObKgUlaWFpfF4oiU29nfc2l468lBkxM/iu+DtSTak0OZt6jViXAQoWQtw4iL1UitWozwUU+I3xlCVNpSFhLFHaMJJsrShc7dBeg4DA27lvB3aRZk4j0ZTISOtI3tv2XiBVF+D0jNN5aNVDPQ75ecHnXDfuOpJMSWxv2U6IKoQIXQR1zjrCtGGkmFJQK3c/TSsz9OizMyCK4r8BBEHQi6I4pMPwC+ptxJj2Pl8VFaqlvNnB6PidnIFlz8CoM0Ctk14zboQvroMbVoNmjxWbZXrD54Etn0qjLV4XaIw0XfAez1cs5MNiKdgswZjA7VNv59n1zzIjdgZjI8fywqYX8It+wnXh3DzxZp4+6umA6uDMuJl8V/wdftGP0+tEpVARrY+mzdXORUv/QaevE51SxwOzH2B+8nzZITgYuNoBEY7+Byx9Amz10hPrzJvweTpZHJ1KfctWnt/4PKnmVG6deCvLa5ZjVBuZGD2R97a9F9iV3WMn1ZyKQlBw88Sbuf332wnThvHYkY+RESZPAwWhNYKnE37tumHHjME/+SoWmUzcveq+7t/DpL8x/5eHUdjrcZ33Jh2eDgBGWkcyLXYar2x6hRhjDH+f8nfW1K6hw93BzLiZuP1uXL6eunIiIqlhqfxS8UtAmnhazDRSzal8nP8xt066lbOzzkavlq+Zw4U+XyUFQZghCEIusK3r/ThBEP5vL5sNCsUNdmLD9h5NHmGUnIEAjmYoWybNu+0gYTJYMyS9b5n+U7sZvrxBcgQA3Da2NuUGHAGASlsln+Z/gkFtQKvS8uH2DwNPJk2dTbyT9w51jjrOzDwTn8/Hk2ufJK85D4AsSxYVHRWckn4Kj695nJnxM7l67NXMjJ/JPX/cQ1n7QauYfXgTlgwT/ySNAE28RHKoFUr442nKRp1ImyDw+tbXaXe3s7FhI8+ufxaXz8U5WedQ3l7OcSnHMSp8FCClhi4uW0yYNoxHVz9KU2cTRW1FvJ/3vqyB3xs+LzRsg+nXSaqDHgdlhlDuXvUgnT4pSLrT18nda/9L2ZxboK0CtcfJuMhxAMxLnsfrW1/HK3qp7Kjk4VUPS/oOORfy7PpnKW0rJVofHThcqDqU68Zeh9/v7+For6xdiUFtQKPU8NiaxwICRjLDg/48Mj0FHAs0AYiiuBGpLsGQQhRFShrtxIX1bWSgtHEnoZP8H6TUG/Uu2068BFa9INULl+kbnW2w6hXpM935Aq7WU+JuCVp9fcMGMi2ZVLRXBC3Lb8mnwdnAv5b/i4tGXkSaWdKonxozlVPST2FD/QbSTGncOOFGam21vLTpJWpttdw44UaaO+X00YOCMRJGny5VzFvymHRzOuZ+fNFjKPY5aHO1Ue+oD6zuFb3Y3Daq7FX8XPEzr295nThjHLdPuZ1jU47lp/KfgkRrllUvw7arkI6MlNI57jzY8D6sfBFS5tDg9wQcgR10+jpp9EtOeVNnC5ePvpwZcTOCnvoBfq/6HYvOwmkZp/FV0VdcNeYqxkaMJTMsk+vGXcenhZ9y19K7iDXEcsXoKwLblbSVBDKDah3y9XI40a/xU1EUd71SD7nInoYOFwoBTL0EwuxKtElHyc7OQMFCiJsQvKIxSsqX/i24FtMPJT9w1+938VPZT/tj9qFH+XL47i+StLBip9koj4NUjSVo9UlRE2lyNPWaQpZlyQJRmk5I0IbzxrFv8P0Z3/PU3McZYU5jZtxM1Eo1r255ldxmSRk7tzmX17a8hlFtHLBTlOlCFKXUtS9vlDIHQJq3/uVBChf8g78tvYtaR21Q356QdgL/Xi7VJPCJPn4q+4ncply+Lf4WpaDs1p3oYmbcTIwauT+DMMXC0ielGA1vJ2z6gHhTEqOso3qsplPqAqJB4ToLr215Da1C22uAbowhhnhjPHdNu4tPTv6EecnzeHHBi9wx9Q7+u+a/gT5bVL6IKltVQPwr1ZwacOJi9DEDe94yB5T+OAMVgiDMBERBENSCIPwVyBsgu/aZgnobida+zVPFmHWUNnU5A6IoBQ/Gjut95VFnwuaPob060PT6ltd5cu2TmLVmHl39KG9sfWM/rT+EKOhyjjZ9BHP+BjsKlmiMjAofyblppwRWTTAmMCdhLpH6SEI1oZyXfV5g+DFcF8752eeTZk7Dog3jqU3Ps6V6BeHqUEJ1YaQZ4nG72ijrKAuUvN1Bg7OBDnfHQTndwxK/T3L6PrlCqjTZvIvMdGcbpe5mfKKPiJAIbp54MyaNpEqoVqhRCsHlwxeVL2Jm3Ezunn43pW2lgfZ0czrn55yPStGfmOdDnLpc+P0pKP4t0NQy8hS+PeYu/rH1RTLC0rh9yu2YteZAzEByXRGYk/Co9WSEZZBiTuGX8l+4ftz1qATpszVpTDw4+0Ei9ZGolWriQ+OJCInAqDFSaavsEcgL8GvFr0yLnca0mGnYPXbcPjd/nfxXOR10mNGfX9a1wNNAPFAFLARuGAij9of8ug7i+zBFABBp1NLi8OB0+whpL5HmOEN3482GhEHGfPjjaTj+EQpbCnl186v8c/o/sYZYGR85nvtX3s+0mGnkhOccuBMarljTpb8tpZITNfPPUiW19HmEh6fzl+hxnJp+Kpta8yluL+GBlQ/g8XtYU7eGeUnzuGbsNfhFP3aPHR9SkRSFoEBAYEn1Ul6c8xgzU48l0pLGXdPuIq+lAAGhx4VKISgkiWOZgaF6PbxxkjQNFJ4qZRL4ehaIMik0nJpxKovKFtHkbOKcrHOIM8Zh1VmDhrEB4gxxZIRl8GPpj9wz/R7mJMzB7XOTZEqSswl2pqUM3jlDGgkYe67UZozm26TRPJL7MgBrWMvC8kU8O/dxomwtJC95BsX8e+GKhegVIk6vk08LPkWtUJMYmsgN428gLSyNbEs28aHxvR42TBsW1BZriOXEtBM5N+tc6px1nJN9jpxNMAzpTzZBI9DvgkSCIIQBrwCjkeSLLxdFcXl/99NX8mr67gwoFAKxZh3FjTZG1a2A6FF73iDnFCkqfu7tPLnuSY5PPR5riKTpbQ2xclrGaTy06iHePO7NPQoeHRZkzKdY4WeL4MEt+hklhDAifjpChBQNHhIShtJu5OHVj/TYrLC1kDMyTufRrpzmjLAMXB4n1427Dq/oRRRFdCodX5ctYkbyAgSFAoM+gtEaI1ePuZIXN78c2NfVo68kxZRy0E75sKPo5+54kNwvYerVsPy57uWjzySrvgh70hziDHE4vU4MGgNfFn3JxoaNvH3MK4wJH83mrvLESkHJmVlnct+K+3D5XFTbq5kYPXEQTmyIIYpQsxFqNoBSC/EToLkUOrqKdIVYIDSWhpEn8XLZdz02dXqdFLaV4taFEzbnZiw168DvwmVNpt5Rz00TbqLD3YFBbaDKVkWqOXW3jgBImQdjI8ayqXETIPXZ7VNvZ2T4SACSzEO2XI3MXtirMyAIwrOwy7jQToii+Oe97OJp4AdRFM8SBEEDDGiuybaadk6fsPsv867EhYVQ3GBnVPlKCN9L2pIhApKm0/z7Y2xq2cQFIy7osXhW3CwWli5kde1qpsZO3RfzDxnyfTauKPlQqkSINCz8atw4do7I2J3DlIGK/426ljZDOFn6WAq97dy/8kFsHil4LEQVwj+m3dVjG53fzyVuFVNzrqJG9BIrqBjhVqKVBWsGkJ0uC40FoI+A4x8FpVrKdy9chL+jmje2bmJj180D4Nqx11Jtq2Z7SyEPJp9K3siLKGorQaPU8G7eu4GANjkltIuKlfDmyd2jLnornPRU9/KlT8D061EkT0O5PrgiZKunnYfXPMoFycdxS8kWQhbdi+ryHxgXOa6H6NC4yHGckHrCHk2JMcbw+JGPs615Gx3uDtLMab3GHMgMP/oyMhBcP7aPCIJgRso4uAxAFEU3MGCF5v1+kcIGGwl9jBkAiDHpKKjrgKo1MPmKvW8w4mQ0P97B3FkXo9mlKphSoeSYlGN4ZfMrh70zsLRqacARAEly9q1t7zImZjKqrviBZFMyx6cez/cl3wfWG28ZQXbhEqzr34dp10LZc3wz6RTmJ88n1hALQKOzkTW1azklcrIUta41gaDE9MOdBH3qybN2Hwcis39EjZTqEXi7hvvLl8Pc2+HXhyHnJChYyPZj72Xjtp8DmxjVRkTEgAxuq6hgNgZ+aCng54ru9abGTJVHdQC8Hlj2v57TL46uDBlLGrQUS2m7S58knNu4YcRF/Gv9k4FVd8RoALxX9gOnZf+JnLJl+Bvy+LH0R64Zew0iIhqFht+rfsfpc+7VpBhDDDEGOTjwUGOvzoAoim/2ZUeCIDwriuJNuzSnAg3A64IgjAPWAjeLomjfaburgasBkpL2b4ipvNmBUavCqO17KER8WAjba1olBTVr6l7X94QlUiH4OMunwdPL8hmxM/i84HPK2stINiX33fghwIHsi1pnQ1BbtbMJr7cz4Azo1Xpum3Qb02Om8Uf5z0wyJHKE04V1SZfCtbMZQuNJDk1mWf5HfFH4BQCpplQuyLkA/ze3oSiQKh9y4pNBxwMkOdZhyIHsiwGjfhscdTc05Emfc/xkcLZKSpOrX4EF/8Gh63bMlYKSG8bfwPMbn6fd3Q7ASSnH89e4o7lj6h0ckXAEy6uXMyVmCrPjZxOmCxuc89qFQe0L0QNtvWhllK+ACz+SKhZWb4CMo8Hn5RhRR/iM+/m2egmhWjPxxnhe2vRSYDMHkn6H32Pn/BHn8+z6ZwOxG2dnnY2Kw3x68zDmQI7D9SZOrQImAs+LojgBsAN37LyCKIoviaI4WRTFyZGRkftlwNbqdlIjDP3aJsESQn5Ni6Shvmv9717Y1LCRzdZ4Mrb1nkqoVqqZGT+Tj7Z/1C87hgIHsi/mxs0Oajsv9UR0OnOPthhDDGfEH8njGedzakgSbYmT2Xbhe9Sd+BhF487BprfQ0tlEfkt+YJuS9hJqOipRNHW3Ub8VwlJ6HjAsBaxpfbLX1umhoL6D6ta9PxkdDPraFy6vj+IGG2VNdvz+3c7mDQxpcwEBsk+EcRfADrW5MedIgaM/3k2qx4NWKTl/RyQcwbcl3wYcAYBvSr8nVyUQa4zlrKyzePzIxzlvxHkkhCYc3HPZA/39XVS2OCist+F0H4ApKrUeplwJ8ZNgzl+lssThGZCxACIyIfsEmHG99H7K5YSOu4Ajs07ljvE3UdCcz1PrnsLhlYTVEgxxJLXXAaC1pvNm7ps9gjg/zv8Ynyj20IPYGx6fh7L2Msrby+UaEsOcgZ6UqwQqRVFc2fX+EyTnYEDYUtVGgqV/ZTNjzSFUtXtxmfsmc7q8ZgUkTUfbUUtIY1Gv68yJn8NXRV/h8fU2dnB4MCFmMo9Ou4eE0AQiQiL4+9jrmdtbIaHKNfD68RR0VHBf+yYuXH4XF628l1f9DTxT/g2PRUaypSk3aLP1jZvwm3aKDVn3Fpz4mFQyV2eW/p737u6zQ3aioK6DK95cw4InlnDiM7/z9cZq3N6hf2GraHFw9+dbmP/Ebxz71BJeXFJEq2PAZuF60l4l9V1HFXx2JXxwvlTps71CGiGYfRsYIsnY+DEvznqIceGjyQrLIr85P2hXNe5hUQx1r9hcXt5ZUcZxT/3O/Cd+45YP1vcUNdtXEqdDRBb8/gSseF4qQxyZLaXtvnkivLoA3j9PGi3oIry1in+MvIyj42YRqg7l6Pg5PJV1MZFbv4Jp12HXW3u96a9v2MCNi29kXV1w7MGu1NpreWT1I5z6xamc9uVpvLDxBZqdssjXcGVAnQFRFGuR9Amyu5rmAcFX9gPEpqpWUsL7NzKgUSmIVjsp1o3c67qdPhdbG7eQYR1Ba9JUorZ+1et60YZoYg2x/Fr5a79sOZTQ68M5fsTZvD//FT4+5g0unnAd1rCeQ6zujlrs698CrYmFnka+LfkWv+jH5XPx/rb3GWkdyU/lP5MTMTpo//MjxqOo3+mr5HNL89dnvgLXr5D+xgRvtwOH24vT48Pu9nL/t3msLJEuYi0OD3/+YD15Nb3rE7i9PmydQ8PJ+2JdFZ+srcQvQqfHzyM/bCevpp12p2Rfu9OD1+ffy176gc8Lzq4b9/YfwGOXbk4eJyRMgZixEJENlWslDYJLvkI44QkmfXYTL9gEzrCMY0rMlKDdWrTBIlTDkU2Vrfzjiy3YXFKGxe+Fjfy2vR67q7s/PPvSH/k/QN5XkHUsJM+U0psb8+GLa8EULzkHrg7pc28plbYxxZL9wWU84gnl89E38Yg/nGxbG1z4KRz7IBGh8aSGpvQ4jICATqUjrzmPB1Y8QFVHVZApO7O4fDEfbv8Qn+jD4/fwwqYXWF23uv/nJzMkOJDOwO4mm24C3hUEYRMwHnjwAB4zgCiK5O7DNAFAIvXkC3uPF9jSuJk4Yzx6VQhtSVOxFv2C0tW7POrMuJl8lv9Zv2051AgzxRMR1jN2wi/6WVO7hj8vv5eL/ZX8PPcmfqn8LWjbwtZCwnRh1NnrODHluED7/IS5zEs8GnYEMSnVUtnW2HGSlLQpLlhSuguby8v3m2s4/6UVXPTyCpZsb6Bll6dpUYTiXZ7o/H6RlcVNXPXWGs58fjnvrSyj0RYs43qwaHN6+Hx998XaoFHyjxNzeGt5GW8sK+XB7/I4/f/+4G8fb2RL1QF48q7ZCF9eD6/Ohw3vSZU9m4tBHy5lEISEw+YPoewPMMVA9vHw9c2w+lWYdTPGop9pDTEwI24G6WGSBoVKoeKCERf0Koc7HNlW2+1AHpkVydVz0vhwTSWXv7GGz9dXcvMH67jzs81sre5Hf7hsUr2HWTdDWxW4bTD/X9BWIX3nQ2NBHwlH3Qm5X8BHl8Dq10BQwlmvo9v8MdGfXYtu/VsQmSU5yAoFFlMCD0y7KxAIGKIK4Zpx1/BDyQ8A5Lfm81PZT3xb/C3trvYgs9w+d6BA0c78VhH8O5YZHhxIOa+ne2sURXEDMPkAHqdXato6EQGroZ/Vlf1e4n3l5LqO5NS9rLqmdm2gappXZ8YemUV4/k/Ujzk9aN1J0ZP4cPuHNDobiQiJ6J9Nhzh5TXlctfAqvKL0BPW/bW+RZclie8v2HuvFGeNYWrWU5dXLeDvtfP6UnYAYlkLi2rcwJNTAZd9AaylojJLIkXLvX+cVRU1c9273EOjad9dxz8kj2VzVhrjTlHvELt+jLdVtXPjKSrxd8/J3fb6FTo+fy2fv3YkcCPRqJaPjzQGn5bJZqTz7cyEjYkLZVNnKojxpCLiowc6v+Q18ccMskvs5ahaguRjeOhWcXTUllvwX0o8CrRmmXQOL7pVGBwAatku1PMqXQ9bxsPRxMETivPx7yh01fFv8LSOtI5mfNB+QFAd3FCga7sSYJPlkvUbJ+KQwnlrUXahnVWkzdxw3goe+38Yv2+r54oZZfVNKVYdAVA78sFOoVcUquORL+PgyqU/OfBU+3SkT6ttb4eRnYNKlcPUSaZ0QC1h6jsyNiZ/Be/NeYmtHKRsaNvJV4VdU2yWFVbPWTIurhcfXPs5/5/yX41KP67GtWqFmTPgYtjRu6dEupxkOX/Y6MiAIwteCIHy1u9eO9URRfGNALd0LmyrbyIg09l/sp7WCJK2T3OY9b+cVvWxu3NyjhGpr8gyiN38u5VTvgk6lY0LUBL4u+rp/9hwGbG7cHHAEQCpENCZiTA91sxRTCmmmNK4YfQX3jbmOyLYasn+4hxE1WzEULJKi1UMsUmBVZHafHAG/X+TtFaVB7RsrWsmJ6U7BOmlsLDlxpqB1vLsE6L20pJimQRodUKsUXDUnDVOIdN4qhUCb08PUVCuLt/WcC25xeCio248CP/Xbuh0BkJwDc6I0IiMoux2BHWz6CEKjIWoETP4TjL+AEq+dso5yLh11KT+W/siLm17kxU0vYtFaDhnZ2vGJYUxLtTIjPZyfcut6LBNFyTGLMelosrsprO+jTLbPDZs/CW5vq5L6JOUIKP4lePn6tyXp9Op1kPu59Lc9uHBQpCWVnPCR1DvrA46AgMBN429CQODWibfyQ8kPdHo7e2wnCAJnZp2JVWcNtCWFJjEnYcjVrpPpI30ZGXhswK04AGyqbCU5fB/0jJqLSDKreLtlz3N525u3Y9FZCN2pUIrTmoaoUGKuWENbUrCuwMw4KavgslGXyYqEO6HrRZP+i4IveGHGfZR42qhzNlLvqOee5ffgF/38c9RVTNkhcLPjc9SaQNG/WS6FQiBMHzxyFG7QcNN548mv7yAsRMOI2FCsBm1Pm9XBNht0SlSKwevXMfFmvrx+Fvn1NlweKeDR5xfRKBW4vD2/zxrVfswIqrTBbUuflNI5d7lJSAfTS+1+Hyx7lvaxZ/Oyr5pFFYuJCIngijFXYFQbseqsJIQmHDJPk3FhITx7/gQqWpw89uO2oOUhGiWursBUtTL4+9QrgkKSQt+VHVkAnW3Qm+S2NhQKFsPXN3a3jT1PCrLV9lw/2hDNnyf8mWOSj6HR0Uiru5V3896lpL0EpaDknhn39FpHItuazTvHv0NhayEKQUGWJYtYY2zfzktmyNEXnYFhMQm0oaKVGen7oF3eVEiExYytXqTNJWLW9n5xX1e3PlA6N4Ag0JIym5gNH/bqDGRZsnB4HWxp3MKYyDH9t+1QwWWD+jwp+ExrYKwmHIvWQour+2nzqtQTGfXuBdSf+l+eWvdUj82fLviQOXEnEROZDfZGqQpi5vw9HrK00U5BfQc6lZLsmFCiuoZwL56ezLebagJP+VqVglPGx5ERHUpGdPdF0u7ykl/XQW17JwlheiYkhmE1aGi2d8cX/O2YbMy9OBcHk9RII1q1kk2VbTx8xhjMIWoeOXMM//o6l1aHFLg2Ks5Edsx+1GiIGgUxY6B2M1hSYOKlYAiXpIjDEsGcBG3l3etPvhw8nZJS3ogTKYkdxaL81wBJMOr5jc8DcPuU25kWO23f7RqCON0+6ts7OX1CAhdOT+aR77dR0eLEoFESHSrVQhmXYCYrpo/VF1VamHUrRI+W/lfpwJoCSp0kn167Ceb+Hda9CR4phRBBIfVR0S4jBps+kKZ14oMTunYICf3fhv8L9A8gVZMs/YmT007u1bxEUyKJpsS+nYvMkKbPMQOCIGQCDwEjgUBtUVEU+5bIPYDsCB68cNo+iPw0FqBImUWySUFek4/pccEfiYjIhob1nJZxWtCy9vjxROT/iKF+G/aonk84giAwM24mnxR8cvg6A55Oqcb6rw/C2W/Cu5eSNv5CXhtxOStcDTT43cwKTWVswW/gtuOyB4sVOTwOvPETwJwsRUzPu1d66hlzbq/TA5sr27jo1ZW0dUXVT0gK45nzJ5Bo0TMhycLH185gSX4DKoWCI7IiGBPfU/ug0+Pl9T9KeGyhlAYnCPDwGWP56JrpLClopKHDxZzMCCYkhR34z6ufFDfYuPLNNYHYgahQLZfPSuHxs8exrLCRrJhQpqeFE23S7WVPe8AUAyc8BiVLpGJT3/1FeuqfcSOsfAGO+Q+0lEtBbTGjpViC1lIo+Q38Xlx6a6+7zbRkEqnfPz2LoURBXQd3f76FVaVSZopBo+TZCyZQUNtBTpyJlSXNPHrWWGakhRMV2o/+0IXChnckRxikabEpV0v6A+01UL0JznlLitNwOyFpmtRXIb1kafj2nHpq9wSnQra6WvGJPtTIhYcOZfoTQPg6cC/wJHAU8CcGXqegT1S1OlEohP4HD4p+6aIVejaJJgW5u3EGKjoqAYgI6eXCpVDSnD6XuDVvUXBCcKLE7PjZ/GvZv7h9yu3o1fswjTHcadgOv9wHM2+B5f+T0tH0VjK+uJmMo+6CDR+DrU6KlgYyvD5CVCE4vd3z0GdnnEr0r49Jc9WCAuwNcPpLvToCxQ02nvulIOAIAKwvb2V1STNapYK69k4ijFpunp+F3+8nt6adPwobSbLqSeoKsCust/P4T9358KII9361he/+fASXzxqcgMHdsSivvkfmQ32Hi8IGO3k1HfxpVjIalXLfAwd3xpoOpUslp87vkwLbVFppLrpyDWz7Tnpfs0EaQYibII0M5P9IyqTLSApNpLyjIrC7keEjybZk7/54w5ANFa0BRwDA7vbxf78W8dJFk2i0u5iSYsWiV/e5xDogfdarX+12BED6TWn08MV1kijRMQ/Ae+dAWLLULyv/D457WCpxPO0aaVrA1QFV6/cqwnVkwpG8lftWj7aLRl6ETrUfzqTMsKA/zkCIKIqLBUEQRFEsA/4lCMJa4J4Bsq3PbK5sIz1yHy547VVSJLpGT0Kohy2NvQvNbKjfQLo5fbe5k22JUwkvfLTX0QGLzkKWJYvvSr7jrKyz+m/jcMfRKN1Nramw/k1pmHNHqpLPDS0l0v9Va2Du7WTk/cjL02/j5eolFHeUclrifE62jkOtipeEhUSfJGgTNz7oUD9sqWVJfj1bq4NTofJrO3htaQlbqtsx6VQ8ee44ato6eei7bdjdPpKseh49cwzT0yNosrt6ZBaAlMe/Y9h9KLGxoiWorbjBRoRRQ127m9s/3cjfjhvBaePjMGj348nOGCkFq/18n/Q+xCoFqIEUzHn0PySnoGYDjD5Dyn9fIlWejGou5akjHuWd/I9ZVbuKIxKO4Lzs84aM3PCBoqYtOH6isN5GebODG99fT2WLk1CtirtOzOHksTEYdX14ePE4pd/GrjiaupY7YN0bUprh1s+kWI0j74CGfBh3Pnx8seRIGCLglOfAGLXHw42NHMtz857jpY0vYffauXzU5cyOD1YTlTn06I8z4BIEQQEUCIJwI1AF9HHia2DZXNVGUn+87R00FoBJkj1NMSt4L7f3i/26+nVMj52+292ISjVNmUeTsOIVtp8SHG85N3Eu7+a9y5mZZx5+gYRhydLTS+FPkHGMNG+p74rt6GyTFAI7aqHgJ0lBLfM4xsVM4fHs0+n0dmLWmiQnYOE/IP1oUCilG5JCIQ2XdrGlqo2/fbwRi0HDrIwIPlxd0cMMi0GDTxS5YGoSpU3Sk/PjP+UHbvrlzQ7++eVW3rx8CokWPSFqJU5Pt3MYFaolLmzoPR1NT4/g281SlLhCgNkZEZw8Lo7SRhufratkTlYU//hiC2kRxn2LqdmZyGyImyiNBnTUdD9lejul/pl6lRSktuVjaNpJndPrJLO5gn/M+Ad2t51QTShKRR8D6IYR6ZEGtCoFR2ZHEaZXsyS/gelpVj5eW0llizTS1eHycvfnm0m26pmZ0YeUY60RRp8NdVt7tpsSpFEy0S9pQNTnSSmd5iRY8qg00vbl9d0jCvZG+OxquGbJHmuwaFVa5ibMZUr0FHyij9DeghNlDkn6M8x/M1L54T8Dk4CLgUsHwqj+srGylZR9EBuiYbskUAMkhSooafPj8fV8JGxxtVLvqCNxL1rprYlT0bZXY6pYG7RsZPhInF4na+r2uQDk8CU8Hc57T3pqzD4BUudKgU2zboatX8CsW8DSdXHSGGHCBWBNQ6vSYtaZpQn7goXSDWf7d5D3tfQ0tPVz8HdHzFe1OOhweSlvdmDRqzkyS5rS0aoU3Do/E1OImvRIIz9urUUUIT3SiEHT0xcuqLdR0+YiNcLAS5dMItokRdGnRuh54aJJxJj7J3V9MHB0erlgahIRBg3/PGkknR4fj/ywjSa7h4zoUJLC9YgilDcfAFnckDA45RkpmFD0S7+f2bdI0wOCIM1fpx8tZXqAJAt95J1SX1WtRa1QE6YLOyQdAYAQtZKHzxxDRbODX7fXc+KYWM6fksS7K8t7rOcXoaLZ0fcdjzkTxp4rfcZKjRQrULZUGg3YEY8RkSX91n65X1IqTJrRPeq2A1c7tFf26ZB6tV52BA4z+jwyIIriaoCu0YE/i6LYx0TZgWVH8OD5U/ahmlh9HmTMA0CrEojWC+S3+BkV0X2x2lC/gTRzeq+pNT1QKGnMPpbE5S+yNeEFyWvfsUhQMC9pHq9sfqVXOdZDGkGQBGqu/kUqvXrq/0l69hoTTLxMmirIORU6W6QRg95qCSRNh227qJ2lHNEjtTAyVItaKeDxibzwWzEz0sO5ZX4mR2RGEBWq5eYPNrKuXBpSX17cRG5NO5fMSOb/fu1+go0x6bDq1QiCwBGZkXx142xa7G4iQ7WEG3tJrxsCqFUKNlS08tR547n67bU4uorjfLC6grlZkczKkEYD9iuAcGdixsAlX0NruaQ5qrPAqLOkVLfwLGnOeuSpkDFfGuJe+7o08jPz5gNz/CGMCPzlo43skKN4ZWkJaqXAqDhT0NRVVH/6IyxJEhE64i/SyFiIRfpMQyzSlExnG+ijpN/VZd9J+g5tVZIOxM71UZQaSa1QRqYX+jwyIAjCZEEQNgObgM2CIGwUBGHSwJnWN+o7XPjFfVAe9HZK0c/m7mI3KWYFW3eJG1hXvzYgn7o3OmLHgujDWhScjTkrbhbbmreR20vRncOC0FgpFSosARKnQXQOhKdJFy5zHFgzJNnVb/8Ki/4FVWsJjOFnHy+lt+3Amg5jzu6x+1GxJu46IYcdaf8ri5swalWsL2uhqrUz4AjsoM3pIcbcfUHWqhTcf/oo2ju9PPrDNu75cgsljXZSIwxD1hEAOHpEFHqNksoWZ8AR2MFv+Q043T4um5nCqDjzbvawD+gtED0S3A5Y8T/Y/JF0J9SGgkoD6fOkVLflz0k3rbQjJYfuEKWy2cGHq8rZUtXGroUjP15bye3HjUCt7J4evHJ2KiP6m+qp1knTNOEZktMcPUoa1TQnSP+HRkpxNJHZULYcNn4IR9/T/VAiKKSMkPC+FWSTOfzoT8zAa8D1oij+DiAIwmykDIOxA2FYX9lRj6Dfc/GNBWCK7VG2ONmsYGO9j3O6YgCdvk4KWwqZn7Sgb/sUBBqzjyV+1es0p82RvPgu1Eo1x6Ucx9PrnubFBS/2z9bDgfI/4O2dZJ1X/B/86QcpJzo8Ay7+DBq2ScPTkSMC0zs7cPlEiutt3Dw/C4/Pj1qpYFttO60ODx0uHxqlAvcuRWJSIwy8dfkUGm1u6TsEnPPi8oBgz1vLy3j7iqkckTl0n6ZSIgzcc/LIXoMmDRolczIjSQrXH3iHpnwFvHVyt8O28nm47HtInAKxY+HKxVIxHZVO6i/DoSnJ3epwc+fnm1lW1MQNRwU/NFj1GkbGmnj3ymmUNTmIMGrIiTUN3JRT3tdSrABIMsbz7pG0IazpUj/0QalT5vCkPzEDvh2OAIAoiksB7x7WPyhsrW4jsZ9liwEp/WmXAjppZgUb6rufrjY3bCIhNAGtsu+jDo6ITPxqHdbiJUHL5ibOpai1iGXVy/pv76FGWxVs/16SWq3ZBGvf7rnc65Iq4+0gNEZ6wkw/OsgRqGpxsKqkmcIGe2BkQCFAVUsno+PNfLu5hquO6Bk0NSczAqfbh0mn5oyJCUxIsvDztvog5b4XlxQP+XLG22s70KkVjE8M69F+w1EZKBXwR2EjP2yppaypD3EDHqekfb/pIyj+VZra2RW/X9KO2DnlwueBLTsV5gpLkqYKUmYfso4AQFGDjd8LGvH5Rbw+kfidRpsEAS6dmUJNqwO720enx4fL66e40c7n66vYUNHa/++WpxMqV0v9U/QL2Ju6l9nqpZiBHdTnSaNsPrfkoKkGVyBLZmjTHzfxN0EQXgTeRxoUPBf4VRCEiQCiKO69APYAsKWqnbR9SSusXi89xexEqllBUaufTq+ITiWwpnZNn6cIAggCTelHErvuPZrTj+yWz0Uq7nFu9rnct+I+Pj/l88M3d7e1HD68BGrWS+9VWjj+Mdj2laRqt4OdtAZ2R1mjnSveXMPFM5JIjzLy+MJufYCLpydj0asprLdx+awUbj8uG7fPj0IQqGxxcvXba4kM1fLcBROYlhpO5y6OAECn2xeUZjjUcLh9hGgUnD81kVPHx1HR7MBq0PJ7QSN1HZ2sKW1ha3U7sWYdb10+lczo3QxRiyJs+hi+vqm7bdJlsOA/UjBg94rdanc701vbIY5vp3mB9CgDl85MocMllceOMGhJsISwqbqduz/vLuhz7KhoPD6RX7bX8+x5EzhpXFxvu+6d3C/g82u63489D45/WIofEP2SEx1k5KA/s8kMA/ozMjAOyEISHvoXkANMAB5nD/ULBEEoFQRhsyAIGwRBOODh9Lk17aT0V1TF44Tmou4o9i60KoH4UCluwOP3sqVpK5lhmf22yR41AoXHSWj1xqBl46PGE2+M57E1w6LkQ9/xeaQnkbJl0NozrQ+/HxoLpWXNJV356Ou7l3tdUmGVjJ0khgUFjDgh8Laj08OmylY2VLTQ5pRU1Oo7OtlU1cblR6SQYNFjNag5a1ICGiX8+5RRjE8MI8lq4LKZydhcXpYVNdFkc/PET/mB1MOGDheLuyr8HTMyml1LDVx5RCraXuoSDBWa7W4yo4zEmUOIDNXS4nDj8fl5enE+y4ubeHt5GfNzogEpD/7nXYoY9dxZCfx4R8+2tW9AxWpptKBhm9SHHdWS+uCujDn8dDTSIozkxIaiUIDd5eOHrbWoFAJGrYrSJjvP/VzI8sLGHtv8uLWO8YlhzM2KpKzJzoqiRlp3KaMdRFulJPr0/d97tm/6AOrypP9DYyQNjp1Rh0ilvWVk9kJ/sgmO2o/jHCWKYuPeV+sfDreXuvZOYvub/12zUZoi6KUAS0aYgrV1PlTK7UTqIzGo92HUQVDQmjKDqM2f0xE/PmjxRTkXcf+K+xkdMbpXieNhh9sh3cwX3i05BYZIOPddSRbV55WeZr66UXLCjNEw7brgfTQXwbx/SvXatSaYfh3ES5WvK1sc/PurrfzUddOenR7ObcdkcfOHG5iYZMHj8/NdV659dnQor102lfu+yWN7nZTwcuq4ONIjjGREGtha0xH0pF/UIFX0G5cYxrtXTufVpSXYXB7+NCuVWelDd4i7sN7Gd5urcXn81LZ38tn6KkQRUsL1/PWYbB76fht+EXw7nXBB3R6SgNwd4O5lKqFihSQgFD9JSg1d/xac/RZc+Ckse1Yafp5+PSQE1+c41IkI1fLc+RP5dnM1erWC1AgDTy0uQBSlmJSLpydT3BBcMdKqVyOK8N+ukaypqRYeOXMsqRG9SLdUrYNvboXMBVLmwK44d5rKGXeuJAi19jVJB2Lq1ZJEtIzMXuhPbYJo4EEgThTF4wVBGAnMEEXx1QGzbi9sr+0gwRKCqp/V6yj7Q8rL7YVsq4Ll1T4UilVkmPc98rY9fhJpix9E5WjGu4s2u0Ft4KYJN/HYmsfQKXVBtcKHHfW5PZ9Y7A1SENOffpAUCL+4tjvFyVYnBVbqzDSNPh233krktu9Rpc+H5Nlw8XRA0SPQ6dftDQFHQK0UmJxq4bGF+VS3dnLWRANP7lQ3fntdB5+vr0Kk+wb45cZqpqeHU9JoY35OFKtKes6DzxsR1bVvBTPSw5mSYsGPiKavleUGAb9f5NO1Us64Qavi03VVgWWlTQ6WFzcxKdlCeZMDu6t7mHj+yF5SN3dgTpSyNup3ErhR7xSPU7VWespUauCL6+HKn+CiTwAhKDDNL/qpc9ShVqiJCBm6DtWBID3KyIXTklha2NSjH0oa7SwraiR1Fw0UrUpBZKiW3/K763CsKmnhqw3V3Dx/l+uS3w8bP5CUHS0p0udfs9OIo1Ij6QvswBApaXWMOUv6nfWi6SCKInWOOpSC8pCqDSGzf/TnLvoG8COwY4IrH7ilD9uJwEJBENYKgnB1v6zbC9tqO/qn8w1SME3FSqkKWC+MCFeyptbLuvqNZFv3XTvdr9Zhix1DxPafel0eZ4zjlom38PCqh3lh4wv4xT2XUB7StPUiZNJUKDkC7VU9c50BV+0mfjr9Cc51buXk6q95YuyxVE+6SNINUGqCbiy/5UuOgEWv5q4TcghRq1hV0oxVr6GqNVgCdmVJMzmxph5t22rbGRFrJiXcwOWzU9CqFISolVx/ZDpTU3s6ayqlYkg7AgAdLg8F9TaKG220OoOHmNeXt3JEZjj/PCmHL9ZXYdAoufOEEUxP671oECAJ2JzxkuSUgRSBPu8eKU1wB9XrpSj11lLJ6VOqg/qr1l7LM+ue4dQvTuWcr8/h2+Jve9SaONRYXtTI9e+tp6AueARgdWkLk5KtTEqWigYlWfX846QcftkePF2zeFt9cEChoxlKu+K2t30tpdQmzZDeW1Lggo+kLIFdUWl6dQQaHY28tOklTvvyNM786kw+yf+EDveQkIyRGWT6E0AYIYriR4Ig3AkgiqJXEIS+hMLOFkWxShCEKOAnQRC2iaIYCLXvchCuBkhK6p9wUG51Owlh/XQGKlZJOe8hveddW3QCepUHgWxM+6nA1ZYwmajcL6kdf06PQMIdJJmSuGvaXbyy+RWWVy/noSMeIs7Yj2CiA8w+94UpPrgtLEXKhxZFqeTwToGBW1OmcNuKfwfev130GVq9lT+HZ/eaIjop2cpPufXctiCLrzdWE6pTMzbBzKbKNmLNwVNEE5MsQRfmtAgjD32XR4vdw1+OzeSM8fEgQE5MKMoheOPfW18YtWpSw/UoFAJGXfDPeGKSBZ8fXv69iFcvnUK0SUtcWMjeU3BjRsMFH4CtAeq2wOdXS9M7geVjoehnKb9d3/sT/3fF3/HqFmnA0Ol1csfvd/DKMa8M23LFe+qL/NoObnp/PY02N7N7kRcel2hmcW4tJp2a5y6YwNKCRp7+qYAbjw4edZybFYlGtct3UW+FxOnS6JvfBz/9U4qtOe99SJgi1YzoB79W/spzG54DwI6dfy//N1EhUcxJnNOv/cgcevRnZMAuCEI40pM+giBMB3qZwOqJKIpVXX/rgc+Bqbssf0kUxcmiKE6OjOzfF3trdRtJ4f10BrZ922s9750J09Xi9k7u3357wWlNQeFxoW/I3+06Fp2Fv0z+C2nmNM795lwWlS3a7+PuK/vcF1EjJWnUHQInOjOc9j+pKEpEpqSepuwqkhNiIa+Xe+9nBZ/R2Nl7WEmiRceDp4/mj8ImjDo109PDOX9qEhaDhupWJ0dldxdfSQ7Xc+7kBOzubudjwcgoEEU6vX6+2FCF1wejE8IYHR82JB0B2HtfKBUCZ09JxKhT4feLnDGh2yGLNes4cWwsLy0p5pIZqSwrqOfRH7ezurQZr68PI1DaUEkQytEkVR/cQdRICI2W5q1Pe176fxfaOtv4uODjoPa1dcEy3cOFPfVFUYONRps0MrO9toMTxnRPw8SadZwyLp4Ot4/jR0ezpbKNj9dWcv1RGRyZHckRmd3Ow+g4E6dP6MWp9nZC5jHddThEsWukTei3I9Dp7eTj/OC++aXil37tR+bQpD8jA7cBXwHpgiD8AUQCewwfFgTBAChEUezo+v8Y4D/7auzOiKJIfp2Na+b0wxloKZWGN3dRr9sZr8+DyFYqO+YArftnpCDQHj+B8PxFOKJ2P+WgEBSckHYCOeE5PLDyAapt1Vwy6pL9O/bBRGuQggIz5knDmmFJ3cVQlGpJVz1mLBQuAo8dSy+pTnGGOEKUvetFqBRK7v5iQyDw7+dt9fzjxBxOGReHUavi3CmJXDMnle11HYTpNawtb+b+U0fT1unB6xPJr+vg3q8l5ce4MB26IZwd0B+yokMJ06vZVtPO+vJWbl2Qhc/vp83ppd3p4eEzxmB3e8mts/Hd5lq+3ljNR9fMYHLKHqYKdsbZAoYoqb6AWi+JdDla4erfJGehF7QqLYnGRCo7ek4dReuDHYdDgTB9dyXIbzbVMDM9nL8ek0WS1cDa8hbu/nwzLq+fRbl1vHLpZE6bEE9apAGNSsmz50+gqMGOz+8nPdLYuzCUUiNpcUy9FtRaEJTQWgYtRcHr7gW1Qk2aOS1IBTXJtA9S7jKHHP1xBtKB44FE4ExgWh+2jwY+7xqaVAHviaL4w5436RuVLU60KgWmkH6UZd30oTTftgcVrsK2IpLNTn4v1+HwCOjV+5dk3h4/gaQVL1Ix89pe5/B2JtWcyu1TbuexNY8Rogrh7OzdOy1DDrVW0q3vDaUKYsdIN5d3TmfsrJtID02mqKMMAJWg4uaJN2PU9F4E8/stNUEZAEsLGml1epieZqXT40OpEPjX17k91rtkRjJry1oC6nwKAf550sj+S1cPYaJCdXy3qYZf8xv4daeAtPiwEO44LptnFhdy/tQkoBa/CN9trum7M5B1nCQ3nPuF9F5jhEu/2q0jAKBT6bh23LWsqVuDxy/FiiQYE5gUPejK5QOCVqXg9AnxfL5eChxcVtTEvJwoXvujhA0VrYH1vH6RdeUt3Lag+6EgTK9hUvJevotKFcy4Dt48uVvHwRgDl37Zb1uVCiUX5FzA4vLFgRiOcF04R8Qf0e99yRx69McZ+Kcoih8LgmABjkLSFngeySnoFVEUi5H0CQ44eTXt/atU2FYhpegc8Zc9rralYQvp5njKzV7W12uZFR8coNYfPMZIvDoTpqoNtCfu/YIYHhLOLRNv4ZHVj5BpyWR81Pj9Ov6QImUWXLmIhPrt/C9jHnk+Gw7RQ4Y5gxHhvQRBdaHYNfkfUCkFLpiayEdrKtGoFCzcWselM1J4Y1lpYJ23V5Tx/pXTaO/00t7pITMqlJFxpqB9DWfyatqDKm2CFKKyvc5GfYcL804Os7KXz3K3RI+EP/0ItRulqPboUVLbXpgQNYH3TniP/NZ8QpQhjLCOINGU2PfjDiP8IrR3evjbMVnoNCqMWiXhRm2QXgWAYl/LlydMhisXSWWMFSopo2DnDIJ+MCZiDO+c8A4FLQWoFCpGWEaQbE7e+4Yyhzz9cQZ2BAueCLwsiuK3giDcv6cNBpKt1W0k9UeGeN3bkjSqeveaBHaPnUpbJROiJ5JpcbOiev+dAYD2uPGEFyzqkzMAEG2I5uKRF/O3JX/ji1O/2Detg6GIQinNQcdNIB7oZYa0V04ZF8cX66sCRWAEAWZnRlDYYOfEsbEISIWq4sN6fh+mJFuIDwtB53Cj1ygJN2pQK/uZhjoE8ftFShrtVLU6aXe6SYnQc+zIaEbGm/H5/dhdPswhKl5dWsqUFAumEOlnrlQIHDd6D6mFvRGRIb36gSAIjAgfsUcH71AhLdJIu9PD4z/l88JFk/h0bQWj48M4f2oSE5IsvLuyjE6PH7VS4Mjs/Ujjix4lvQ4AWZYssiy9p1bLHL70xxmo6pIjXgA8IgiClv4FIB5QNla2MT4hrG8rNxdKkdG7qnPtwtbGLSQYE1ArVOSEu3lpoxm/2Narl98fOuLGkfLbEwheF2IvQke9MSFqAhsbNvLk2if5x/R/7J8Bw5wZqVZevXQKX6yvQgTm50gBg88sLqTZLgVvnTQ2lvgwHbfMz2RdeQsTkyzMTA/nxSXFvNNVT95q0PD6ZVMYt4uG/3Dj5+313PDuOlxePwoBnjhnHFq1gid/kgJVI0O1/PvkkZw9OQFRBKfbx3lTEjlrUgLjEy2DbP2hhdWg4bGzx/HLtgY6PT7UKiWP79QPD54+hmVFjUxNDSfKeJjKj8sMC/pzMz8HSWfgWFEUWwEr8LeBMKovbK1qDxLz2C1r34S0uXsp1CGyoWEjqWHSfGhEiJ8QlUhBSz9iEnaDV2emMywRS2n/ChSdlXkWP5b+eEiUPW6yuSiqt+1ddrUXtBoVR42I4tYFmXh8fnJr2nljWWnAEQApeMvu9vHp2kqijFo+XVvJ6pLmgCMAknTvfd/mYnN5ejvMsKCyxcFfP94YKKjkF6Gu3cVXG2sC6zR0uHhvVTmljXZ+3FLLMaNiePjMsUxOsfZvmkCmTySHG7hsVgpev8g3m3r2w1cbqzllbCzPLC5gefEBF2GVkTlg9NkZEEXRIYriZ6IoFnS9rxFFceHAmbZ76ts7cXl9RIb24Sm7IQ+aSyF+yh5XK2krQxAEIndSSxthdbOs6sB48+3xE4jY1r/YSaPGyGkZp3H/ivsRh3q1nD2wsriJM55fxrwnfuOCl1eyqbJ1n/ZTWG/j+y2S7PDmquCs1qpWJxUtTj5ZV0VFixNnL4WHNla00uYYvs5As91N6072nzIuFqc7WO5jc1U7U5ItPHXeeKJN8hPpwaCqNVhYaX15K79sb+TYUTG0OYfv907m0GdYTqCuK28lMzp07wIqAGvfkkYF9lLHe1XtKjLNPYsS5YS7WX6AnIGOmNEY6rehtvXv6WB2/Gxsbhs/lB6QJIyDTmlXVcGyJikSOremnWvfXkt9e99jMewuLz6/SEq4njmZEWyr6Qgouu3MrvEAIb2kEM5Kj8AyjLMJIo3agBOsUSpIjzT2GgcxMy0co061bxU9ZfaJpF7UUCclW9hQ0cqrS0uw6Ifv907m0GdYOgNrypr7dpGr2yrJ4cbvOXCvzl5Hg6OOZHNKj/Ykk5dWl4Jq2/7npYsqLR1x44jY9l2/tlMICs7JPocn1jxBp3f/gxkPNuXNDmyunroC1W2dVLbsXZ62qsXBi78V8bdPNvLxmgr+8eVWHG4f09OsXDYzhdQuwSmVQuDPR2cwMSkMU5can0mnYlyimdsWZKHqGhpPjzRw+/Ej0Gv6EyoztIgNC+HZ8yZgNWhIjzLQbHcRb9Fx0fSkQGxLVpSRS2clMzLOjNXQtxgVmf1nbLyZK2endvdDtJETx8SyvivFcOdUQxmZocawvCquLG7m1PF9kO1d/xakHbnX/P4llUvItuagFHr6RgoBRka4+aNSx9kjeqnm1k9ak6YTv+YNaidcgLiXkYqdybZmk2hK5PWtr3PduF4q/g1hLPrgmAutStEj3a03XF4fzywu4KO1ldx5fA53fLY5sKy2vZMzJyQwJTWck8bFIQjw6/Z6jh8Tyzc3zabB5ibSqCEp3MC0lHCOGx2DrdNLcri+d2GXYcb09HC+vnE2jTYXJY12bvlwA2Pizdx0dCYiImPjzFS3Ojlj4qGZzjdUqWp1srasJdAPVS1O8us7iDRqabC5yIzeP3lzGZmBZNg5Ax2dHgrrbWRG7eWHVb0ebPUw9vw9rlbSVkqjs4HJMb3HFIwKd/NrRcgBcQZc5jg8+nAsxb/RnDmvX9uenXU296+4n1PSTyHe2NekvMEnPcrIDUdl8L9fCgNtd5+Qs1eNiKoWJx+vrSQrKpT15S09lh0zMprX/iihY5cRh2017Zw+MYGk8O59q1UKsg7Bi7DFoKbN6eb3AkloaHNVWyCOYlZ6OOMTzTR0uPoWVyOzX9S1d7KxooXiBgfrK1oDIwEA0SYtR+dEsaakmVm91C6QkRkqDLtpgj8Km8iOCUWj2oPpoh/WvCrJ4+6hvLHL18kPJT8wIXpS0KjADtItHhocSio7DoyEbXPaHOLWvivZ2A8iQiJYkLyAfy/797AKJtRrVFwzJ40Pr57OM+dN4LPrZ3L25IS9RrUrBQG1UoHL6yNsl1EEt9ffq6Sw6hDQEOgrP26p5V9fb+31d6DTKPH4/IHpEZmBo6PTw0Pf5fHn9zf0+p0OUSs5bVwcb105te/ZTzIyg8Cwu3r+sKWGcXvTF8hfCIgQvRt5XMAn+viq8GtijFHEGWJ3u55SgPFRbn4s6WdBpN1gjxqBqFASnt//gkTHphxLrb2WT/I/OSC2HCxMIWqmpYVzyvg4JiZZCOnDnH2CVc8NR6Xj9YtMSbWiVnZfaH/MreOGo3oqsEWGahkd33slykON+vZOHvguj1UlLczJjESzkxOkEOD0CfGMTbAM60DJ4UJhvY0vNlTT6fUTF6brUasA4Nq56czIiCC+v9VVZWQOMsNqmsDu8rI4r55Hzhq7h5UapPrrky7rtWwwgM1t49vib/CKPqbEzNrrcafFOnlxo5kLR9rQqfbzqVwQqB95EokrXqY1eRo+Xd9vYCqFiivGXMGjqx9lZPhIRkUcGEWyoYhSIXDR9GQmJ1v5akMVfzkmm/ImBz5RJC3CQKvDzRt/msIv2+pJtOqZmxV52Dx5uX1+2p3SFMkziwt4+rzxUkVCv8gRGRFYDRpGxB5asstDFddO6av3frWVh04fw5bqNhptbuZmRTI5OWzwjJOR6QfDyhn4aE0FI+NMu0/R8bnh1wcheYZUYQ3w+X20udpoc7fT2tlCeUc5JW0lZFoyGR0+uk/piRF6PylmL18WGDg3x7bf59FpSaY9bhxpix+i4PgH9hrguDNxxjguHXUpNyy+gZePeZlMS+beNxqmWA1aZmZocfv8XPb6aqJCtSgEgQ9XV/DCRZM4MjuKI3cqX3y4EGPScfGMJF5dWkpebQfXvbuOcfFmnjx3HGl7i6WROaCkRhhItIZQ0eyk0ebmunfXMT8nksfPHodZL8dryAwfho0z0NDh4tmfC/nrMbspBezugF8ewqfUst0USVnJ91TaqmhztWFQ6TGoDRjURqw6CyeknYhO2b8f6rEpdp7fEMb0+E6STcElePt9PiOOJ2H166Qvup+So/6OX933OgsToibg8rm4/MfLuXPqnRyfenzfNBeGKVNSrPzvggk8vbgAn1/kiXPGMTO9j5X3DkFUSgWXz0rDHKLhw9UVJIfruWV+FqmRvVd9lBk4ok06Xr54Mi/+VsTy4mbmZkVy5RGpsiMgM+wQhlIw2uTJk8U1a9b0aBNFkfUVrdz+ySbGJ4ZxxsSEHUtwuG20NeThKVtKZMkfFGlDWKxTEWmIIUofRaQ+ErPGjGI3wYH9ZV2dloWlem6Y0MbU2E40+xlTKPg8RG/+jJDmUurGnUVr0jRcptg+jxSUtJXwVu5b6FQ6Ts84ncnRk0k2JaNT7VYoqc8eQ299Mdi0Oz2IiJhDhv1ceL88tz31RZPNRYhGOay1EwaZA9IXbq+Pjk4vphD1IVEMa5A4dJ9ohgFDyhkQBKEBKNvx3jT19AjLUVcE6msK6iZAxOAXCfcHS7B6DsJ3qVqMDvxvoY0vQ27c7w8wSfQFGf6JSue7Wx+21yEIQRBQW9W9PoZ0bOhoKXuqrHinpkZRFI/ri0279kU/iACGswj7wbC/z/0A+9UXfWUw+uxgH3N3xzsQfTHcv/O742CfV7/6QubAMqScgYFEEIQ1oihOHmw7dsdQt6+vDPfzGO727wuDcc4H+5gDebxD9TtzqJ6XTO/I41kyMjIyMjKHObIzICMjIyMjc5hzODkDLw22AXthqNvXV4b7eQx3+/eFwTjng33MgTzeofqdOVTPS6YXDpuYARkZGRkZGZneOZxGBmRkZGRkZGR6QXYGZGRkZGRkDnNkZ0BGRkZGRuYwZ0g5A8cdd5wIDOqr3ekRixtsYmOHa9BtGYDXsOqLQ/jVL/rTF/XtnWJxg020uTyDfY7D5dUv9vV30dF1XWno6Bzs8x3KL5lBZEhpmDY2Dq6IV251G//4YgvryltJsITw8BljmJURcUjr/u+Owe4LmW760hc+v8hv+fXc9dkWats7mZ5q5d+njiI7Rq5eeCDZl99FXk07//xiC2vKWogPC+GhM8YwOyMCheLwu67IDF2G1MjAYNJsd3HzBxtYV94KQGWLkyveXENh/f5XKZSRGWjy6zq46q211LZ3ArCipJm7P99Ch9MzyJYd3rQ43Nz20UbWlLUAUNXq5Mo311DYIF9XZIYWsjPQRU1rJwW73PhdXj9lzY5BskhmV9w+Nw6P3B+9Udpox+fvOdK6pqwl4BzIDA61rZ3k1bT3aHP7/JQ1yd9jmaGF7Ax0YdSpMPRShtCiVw+CNTK70uRs4qTPT2LBJwsobive+waHGWG9fE+tBg0G7ZCaCTzsMOpUhPbSB731l4zMYCI7A10kWfX8+9TRPdoumZ5MZnToIFkkszPPbXiOsRFjOT71eJ5a+9RgmzPkGBFj4vQJ8YH3ggD3nTqauLCQQbRKJtGq5z+njurRdsHUJLLl64rMEEN+bOhCEAROHhtLZpSRsmY7UaE6RsSEYtLJHvyg4vPg+eB8zqz4ncYzXkSht/L3JX+npbMFi84y2NYNGSwGDfeclMNZkxJosrtIjTDKN5whwgljYkmLlK4rkUYtObEmTCHydUVmaHHYOwN2l5f8ug4abS4SwvSMjDMxLjFssM2S2cGa17E1F+LTmsje+i2VM64mx5rD0qqlnJx+8mBbN2RosrnIr7PR6fExJj6M1AjDYJt0WNJkc1FQb8Pu8pIWaSQ1woBWrWRcYph8XZEZ0hzWzoDD7eWV34t5clEBAEqFwDPnjefEsXGDbJkMAKIIq17k57BI9KHjGLl9IZXTryTHmsOy6mWyM9BFbZuTOz/fzC/bGgAI1ap4+8qpjE+UR04OJrVtndz5+aZAP5h0Kt66Qu4HmeHBYR0zUFDXEXAEQMrVvvOzzVTIGQRDg/o8RFcHiztriYqZgE8dgqEhn0xLJuvq1w22dUOGjZVtgRsQQIfLy6M/bMfu9g6iVYcfGypaevRDe6fcDzLDh8POGbC7vVQ0O2h1uGm0uYOWt3d6aXUEt8sMAvk/0BaVjVlnxqA24AhPJ7RqI3HGOFo7W2ntbB1sC4cE1a3OoLat1e3YnPJN6GDR5vRQ3stDxNbqdmydcj/IDH0OK2dge20717+zjiMe/YVzX1qBQaNEreypApZgCSHGrBskC2V6ULiIAp2eRGMiAE5LMsbazSgEBanmVHKbcgfZwKFBVi+BgnMyI3G4ZcGhg8GGihYueW0lHl+wou5xo2KwGjSDYJWMTP84bJyBHUpgv+VLw3jbazu4+4stPHXu+ICWQJI1hGfPn0BkqOwMDDpeN1SvY6VoI94opcx1hiViaJCmdRKMCeQ15w2mhUOGMfFm/n5sNlqV9HOemGQhM9rIXV9swdYpOwQDSUWzg8vfWMPGijaWFTVyxezUQD/MSLNy9dw01MrD5jIrM4w5bAIIq1qcbK3uqQRW1GDHpFPxzU1H0OpwE2XSERmqHSQLZXpQswHRlEBeRyXTk+cD4NGHo/Q4UXa2ER8aLzsDXZhC1IyMCeWK2amolAJ5NR08uSgfUYSatk4y5fTYAaO82UGzXZpW/KOwierWTq6YncrszAjGxJsJlT97mWHCQXEGBEFQAmuAKlEUTzoYx9wVg1ZJiFqJ0+Pr0R6iURFvCSHe0i3Osr22nV+3N1DT1snRI6KYlGyRldwONhWrcIQlovVVYlQbpTZBoNMUi76phARjAr9W/DqYFg46Lo+P9RWtLM6rIzs6lP/7tajHcnOIGr1G/t4OJLvqkJQ02nlxSTGZUUb8fpEJSfK1Q2Z4cLDGr24GBvUxLtlq4O4Tc3q0nT05IUhhsLDexvkvr+Sh77fxxrJSLnltFT/l1h1MU2UAyldQFWIk1tAzzdMdGk1ISymxxliqbFV4/YdvcNaK4ibOe2kFL/9ewi/bG5idEdFj+b9OHtnDyZU58KRHGbhmTlqPtgunJfHS78Vc9OoqftlWP0iWycj0jwF3WQVBSABOBB4Abhvo4+0OhULgjInxjIgJpazJQVSolpHxJsy7KIFtrmoNDPvt4LGF25mTGYHVKE8hHDSq1pKbMp7okKgezW5DJCHNpWiVWsK0YVR2VJJiThkcGwcRp9vHc78UBt5/u7mGU8bF8d+zxqIQBNIiDYyMlcsXDzR6jYrrj8xgblYk+XU2bC4vfxQ1klfTAUjXjtmZEYTp5SBCmaFNv5wBQRBmAik7byeK4lt72ewp4O9Ar9qogiBcDVwNkJSU1B9z+o1eo2JyipXJKVYA8mraWF7YhEGrJCfGRExYCB5vcESw0+3DKwa3H2oczL7YI7Z6cNvY2FnPzPDMHovcxkjM5asAiDXEUtpeekg6A3vrC78oYndLU15nTUogwRKCKIJWpeDkcXEUN9hZXtyE1aAhI8ooTxfsB3vrC7NezcyMCEoa7Ty2cHuPZU6PD4/Px/ryFsqbHUQatYyKD8UcIj9YyAwt+nyFEAThbSAd2ADsmHgXgd06A4IgnATUi6K4VhCEI3tbRxTFl4CXACZPnnzQ7rgrihq5/r31gVGAY0ZGc8cJIxgVb0KrUuDy+gPrXjs3najDIMNgsPoiiKp1+CMyqbbXEKWP7rHIbYhE11YFQLQ+muK2Yo5MPLLHOn6/iF8UUQ3jKO699YVBq+LqI9LIq2lnaWEjn6ytBCA90khkqJbLXl8d+A7fPC+Tq+akYZTnrveJvv4uxiaa0SgVuH3d146b52WyqqSF2z7aGOiPG45M58ojUrEYZIdAZujQn6vDZGCkKPbrEXkWcIogCCcAOsAkCMI7oihe1B8jDzQtdjdPLS7oMR2wMLeOE8bEcOr4eN67ahqvLCmhvMXBxdOTmT8yeg97kzngVK/DZozE5HKgVfYcXvWEhKHqbEfwuog2RFPc2rOc8dqyFq57Zy1Oj4+nzh3PvJxDt++OGhFJdWvPLJmiBhtfb6whwqilqkuM6OnFBczNimRisiyLO5CMjjPz3lXTePn3EipbHFwyI5kx8WYufnVVj4eL//1axPS0cI7IihxEa2VketIfZ2ALEAPU9HUDURTvBO4E6BoZ+OtgOgL5tdJFU69RkrtLmiFAVUsnbp+fOHMID581Bo1KykA4mLh9bpo7mzGqjejVehocDWiUmsOrQl/laqoNYUSpermRK5S4DeHo2quJ1kezsWFjYFFDh4ur3lrDFbNSMYWo+MtHG1l46xyiTIfmqI45RNPje6xWCszPiWZqioUjMiOobe9ke207WpWSzl2yaIYrXr+XRmcjepUek9ZEa2crLp+LSH0kCmFwR4IEQWByipXUcAMOr494cwibq1q5dUEWS/LrSbQaqGh28PO2euraOwfV1r3R7GzGK3qJ0nfH7Di9Tto62zBrzYSo5cDUQ429OgOCIHyNNB0QCuQKgrAKcO1YLoriKQNn3oGhuKGDH7fW8drSUgAun53Co2eO4dp31wfWUQgwPimMv360kUV59YyKM3HXCTkH9WmqpK2Elza9xOLyxVw+6nKcPicfb/+YMF0Yf5n0F2bHz0arOsSHFkURajayPWsOkZren5w8hnC0bdVEx+ZQ0VERaH/ouzxmpYcH+mxWZgTP/1bEvSeP6nU/hwJzsiP5ZnMNkUYtN83LQK1U8MGaCjZUtDI1xcr5U5P499e5/LC1lr8ck83xo2IIHablcys6Knhr61t8VfQVCcYErht/He/mvsu2lm2cm30u52WfR4wxZtDsc3t9LMlv5IHvcmnscHPmpHjUSgUxZh0qpYL3VpaTGW3k7hNzhmxVSbvHzs/lP/P0uqdx+Vz8afSfODXtVBo7G3l63dOsrl3NxOiJ3DLxFnLCc/a+Q5lhQ19c6ceAx4F/AacBD3a93/HqE6Io/jpYGgMrS1p45IftNNhcNNhcPPLDdlocXqYkhwGgUgg8cNpoHlu4na831eD0+FhT1sIlr62ipNF+UGy0uW3ct/w+vin+BrPWTFNnE69teY0OTwcVHRXc8ustbGnaclBsGVTapLnvvM56ovVRva7iCbGiba8hTBuG3WPH7rFT1GDj5231nDYhPrDecaNi+HRt5SHzVNwbR2REcNakBC6YlkRJo51nFheworiZTo+fJQWN3P9tHvNzoqhrd/H3TzaxurR5sE3eJzw+Dy9vepkPtn+Aw+sgvzWfv/32N2bEz8DmsfHqllf5tOBT+jeLeWDZVNnGVW+voaTRQYfLyxvLynB7/awqbuK7zbU4PT42Vbbx2I/bMQ5RMaIN9Ru4a+ld1DnqaHW18uTaJ1lUvoi7fr+L36t+p9PXybLqZdz4843U2eWU60OJvToDoij+Jorib8AJO/7fuW3gTdx/vt8SPLOxeHsdr102hXevnMZH18xgdLyZ9eWtPdaxubwUN9gOio019hpW160GYHrsdH6u+DlonS2Nh4EzUL0OMTyTCntVjyHKnfHoLWjba1AICqL10VR0VPD8L0UcMyq6R9R8hFFLWqSRxXmHbq53bFgI9506mpQIPUatima7m0ijFqGr5EZVqxPLTtr4P26tHSRL9486Rx1fFX3Vo80revH4uuWWP87/mCZn08E2LcCW6nZEUZqGDO/6zL/ZVEPELsHHdrdvyFZG/a3yt6C2Tws+DRpxqXfUU95efrDMkjkI9CdmYAFw+y5tx/fSNuSID+ue39KqFFzTpRf+4u/FpEeGogBSIvRBWQQARt3BicDWqXSEqEJwep20ulqJCImg3tHzJnZYxA1UrsVpjkPRVode3ftQqkdvxVgrOUZR+ii21Jbxw1Z44pxxQetOTrHwzaZqThwbO6BmDyYhGiWRRi0WvYar5qTR2OEiwaJndWkzSwoaUCq6i3ElWPWDaOm+o1PpCA8JD/pNKBXdMT0xhphBnUYLN6i5ZX4mnR4ftk4vCRY9m6tae61aeLCuK/0lbheRL5DqgNQ5gkcBDJqhOdUhs2/sdWRAEITrBEHYDGQLgrBpp1cJsGngTdx/ThwTS2hXWtWVR6TxwaoKHl+Yz3M/F3HrhxuoanXy8PfbuW1BVo/tjh8TQ3YvFeEGggRjArdOuhWApVVLOSntJFRC9wUjMTSRcZHBN7tDjqrV1OlCidRHIexmFY/eirZdesKNCIng6w0tTEu19qoDPynJwu8Fjbh3cfIONSx6DY98v43nfi7kg9UVPLZwOxlRRq46IpUlXcW5LHo1C4ZpdkVESAR3TL2jR9sIywganNK5KQUlN0+8mVDNwfm99kZcWAjvrijnhd+KeWdlOQ//sI2jsqPI2UX86YjMCOKGaGXU2fGzseqsgfdapZaLR17MxOiJPda7MOdCUs2pB9s8mQGkL+7pe8D3wEPAzr/GDlEUh8UE5B+FjTx85hjqOlyoFAL1Ha4eyz9aU8GEJAtpkQbev2oaJY12ok06xiSYD5pymCAInJJ+ChlhGZS1lxFniOO1Y1+jtL0UvUrPyPCRJJoSD4otg4bfBzUbKYgbQaQuYrerefRWNLYGEEXCQyL5Jl/LPSf0fpML02uINetYU9bMzPTd73O4U1BvI6+2o0fbuyvLeOfyqaRHGrlkRjKj481kRA3ezXJ/mZswl7ePf5ui1iLMWjOJoYmUtpcyKnwUWZYsRlhHDKp9W6raabD1vLY8/1sRJ42J5W/HZuNw+9CqFJQ22alocZIUPvSerDMsGbx53JvkNuXi9rsZYR3BCOsI0sxpzIibQVVHFXHGOEaGjyREJWcUHEr0xRlQAu3ADbsuEATBOlQcgma7i6IGOwKS8IrFoKGmzUlJg514i553VpazqqSZa+emB23b0enFoFHS0OFiwcgYZgzgTaOsvYwaew0WrYU0cxpqZffTrEFtYErMFKbETAm0TYieMGC2DDkatkGIleLOBsJDdt8HfpUOUalC1dlGR0sCgsJJyh4urKPjzfye33jIOAOiKFLa5KC6xUFEqI7UCAMeX/DIh8vrRxQEzp0yiGqS+0mnt5OSthLa3e0kGBMYHzWe8VHjA8uzrdmDZ1wXXp+f0kZ7r4Gq7Z1e6m0u3l9VgSBIyTIAJw3haasUc0qQqqdZZ2Zm3Mw9blfRUUGVrYowbRipptRDP/PpEKMvzsBapNRCAUgCWrr+DwPKgUEfKyprtHPbRxtY2xUAODXFwu3Hj+CGd9dT25XPe9LYWI4ZGY1Bo0SlEPD6u6OOTx4Xx+K8Os6ZMrBP3iuqV3DLr7dg99hRCkr+MukvnJ19NjrV0BwyPOhUrILIbKo6qsi27Pki79GHo+2oZUuJFbVxLXDsbtcdHW/ms3WV3H784D45Hih+y2/g+nfX4XD7UCkEnj1vPAChWhUdru756WNGRpMROfSePvtKu7ud17e8zqubX0VExKK18Ny85xgbOXawTQvg9fn5emM1t3+6mVvmZwZdWy6enkRFsyT+tMMRsOjVpEcaB8PcAWNN7Rr+/POf6fB0oBAU3DThJi4YcQF69fCMUTkc6Us2QaooimnAIuBkURQjRFEMB04CFg60gX3hx9zagCMAoFIqeHpRQcARACmqd2KyhRXFjfzvwgnMTA8nNcLAtXPTsejV/PescYyKMw+YjfWOeu7+427sHilV0Sf6eHTNoxS2Fu5ly8OIsmX4IzKpd9QRoQvf46oevQV3cyPbq8CrzcXr9+x23cwoI8UNdtqcu19nuFDV4uTWDzfg6KpL4PWLFDba+c/XudyyIJMjsyJJtIZw6YwUbjgqY1hL3m5r2sYrm19BRLqLtrhaeGDlA7S7gwXDBouSRjt//3QTbp+f91aVc9cJOUxOtpAaYeDqOWkIgsD1R6Vx3ZHpJFpDOHlsLG9fMY3kIThFsK80OZv45x//pMMjTVP5RT9Pr3uagpaCQbZMpj/0J6R1uiiKV+14I4ri94IgPDoANvWb37oCpHaQFR0a0GpXKODSGSlkRYUSqlNy8tg4Ei16nr9QCogRkIY93D4/flFEuduwtf2j2dkcFAkN9Np22FK+nMapV2DUhPaYPukNj87MylI/qVFq2kNCaHQ2EmPofehVrVSQFW1kTWnzsJcnbrK7aHF0OzValYJOj4+jc6LIrW5HROSUsfGoVWDSDs1c9r7SWwR7blMuba42TJqhUZGxrr0Tj09yVipbnDzwXR4z08P5z7xRfLC6HKteg9cHfz82m6vnpGHQKNGoujMg2hxufKKIdRg7ba2uViptlUHttY5axnEYBD0fIvRHv7NaEIR/CIKQ0vW6G6geKMP6w9Ejel7gc2vamZZqJdqk5cWLJpFf28FjC7ezKK+eKJOOS19dxc0fbqCixUFZs4NbP9zAcU/9zj+/2EJR/cDoCoSHhBNj6JmrKyAEtR22tFWCq4MyQSRyD/ECO/CGhPFLtZmcODUWbRgNjoY9rp8VE8qyosHLQT9QRBi1RBi7g1rPnpRAXFgImyrbWFfewnlTk9he185by8t57KftbKsdOk/R/aW3NLexEWMJ04YdfGN2Q4xZh1bVfRn1+UU2lLdS197JiBgTi7fVc+076/h8XRUKQQw4AnaXl683VnPa/y3jlOf+4L2VZbQ43Ls7zJDGorWQYkoJao/djXMuMzTpjzNwPhAJfN71iupqG3SOGRnNnMzuG4hBo+SmozP418mj+Nsnm/ijqIkmu5svNlTz9M8FXDIzhV+3N7Awt45LXlvFz9sbaLK7+WB1BXd8tpn2ARhOjtRH8tDshwIXMrVCzT9n/JN0c3BA42FJye8QO5Yqe3WP1Kbd0aaOYGNHOFmxasxacyDFbHfkxJhYXjz8nYG4sBCePm8CphAVKoUkoX3351soabQzKz2CB7/LY1FePc12N99squG6d9bS0DG0dfB3R7Y1mz9P+DNKQbqBRuujuWvaXYOaPrgrqRFGnjx3fKCGiVGr4rYFWZS3OHj8p3yq2zoparBx28cbWVncEthuTVkLN72/npJGO5UtTu76fEsgBXS4YQ2xct+s+wjvmtpTKVTcMfUOsixZe9lSZijR52mCrqyBmwfQln0m0arnuQsmUtJoA0EgNdyAKURNyfoqWh09b+zrylq5bEZKQKGtZZflq0ubqWhxMCrkwMcPTI6ZzIcnfUitvZYwbRhJpiRUiqEpPnLQKVoM0aOpsBURb4jf6+rLHclkqmrRqaMwa83U7WW6JT3SSEmDHZvLO+xL+c7KiODbm47A6fby9OLueVmrURMIVttBSaOD0iYHkcOwBLdRY+SyUZdxZOKRdLg7iA+NJ1o/tKZ5lAqB40fHMCImlPXlrRTUd7C1pp2t1W1B6y4paOCYUdJI4Pebg1VR315exvFjYtAoD25xtAPB+KjxvH/S+9TYajBpTSSbklErhvc01eFGXwoVPSWK4i07FSzqwVApVFTZ4mBLVTtqJXi8fnJr2rH0ohGgUSoIN2p58PTRhPYyp6pWCuh2mtOjvQaq1kBzMe0Jk9mi8JPXmk+YNgyTxkS4PpzR4aPRKPumRxBnjCPOGDz8eVjj90Pxr3DsQ1Rv/J2xEXufZ1zWGsVY1gJjMWvDKG4t2uP6GpWC9EgDa8tamHsIlI5N7FIS3PkmrxR6j3dp6HDx/qpyJiaFkR1zEObafV6o2QBVa0FjgISpENn7U2J5ezkbGzbS3NlMUmgSTZ1NhOvCGRsxlnB9OGqlmkxL5sDbvB8IgkBapJE/ChtRCAKTksK6qhL21H2w6jV8sb6KsQlmonuppBkbpkM5UJUXm4qhcjU4miB+IsROAPW+xSmUt5ezqXETLZ0tpJpSaXW10unrZFzkuCBxIpnhQ18ekd7u+vvYQBqyP2ysaOXcl5bT6fHzzHnjufkDSVXwmXPHMzM9vMdc8Z9mpfDO8lJOGR/PN5tqmJEezvKdlt90dAbJ4V3pMPYm+PY22P4dvohMPhZsPFXwQWDdiVETiTZEc1zKcRyddPRBO99DjtqNoA7Ba4igydlEeMieMwlEEdY2GPibUITL4yRMG7bXaQKQAktXFjcdEs7ADo4bHcMnayuxubysKGnmxDGxfLvTU+fZkxJYXtTE2yvKCNWq+OCa6QOaNQNA2R/wzumSiBSAIRIu/QaieqZ2VrRXcN2i6yjv6Na4v3Xirfx39X9ZkLyAO6feOawkb0fEmHj+1yJq2jv554kjWV7UFEgzDNOrUasU3PLhBqamWrnjuGxeW1oSSAXVqhRcNjOlh3T0AaO5BN45A1pKutvOexdG9L9uXGVHJdcvup6yjrJA2y0Tb+HlzS8jIPDaca+RY5WrGQ5H9uoMiKK4dqd1l4mi6NzT+oPBJ2sr6PT4USmgye6mqlUysaTJzhkT4jlpbCxVrU5Sww0UNdj5fmsdIVoVuTXtTEmxMn2+lVanhzHxZo4eEYVK2eWdN2yH7d8BUDHmDP6v6NMex11Xv47rx13PE2ueYGL0xCEV2DSsyPsWEqZQ65CmT1TCnodJa+xKPH6B6BAv1c5WwgxWmpzNdMth9E52TOiwLdSzO6anhfP8hRPZVttBk91FVnQoo+JMNNhchBs0jIwzsalSGrLucHlZuLVuYJ0BtwN+fbjbEQCwN0gOwi7OQG5zbg9HAODjgo+ZnzyfL4u+5NwR5zImYszA2XqAya1pp7pNis945fdibluQhc3lxaBV4vWJPP+rNHq1qqSZZrubj6+dwbryFrw+kYnJYQPXL9XrezoCAAv/CUkzQb/3+JydyW3K7eEIgFQgakHyAr4o/IKfy3+WnYFhSn8mTy8BnhcEoRn4HVgCLBVFsWXPmw08la3SD1CjUgTyrwF8osjCrbUckRWJVqVge52Nj9ZUANBkc2MOUfPRmgoUAoSolcSYdD3lhz3dlcVcShVuf3C0r1/00+JqweV1gRbym/Op7gqCG2EZgUZ1cOSMhzV5X8GUK6i2VRPRh0yCzQ0a0swe/IIRtaMZrSkWtVJFm6sN8x4csqzoUJ5aVECnx4dOPfzmZXfH1up2Eiw6nB4fpU126tpd/LilhrZOL8+cNx7/TmV969oHOJjQ5wZbLw5XRy3YGsHY3b8OT3DlvtbOVmbHzybOEIfb6yavKQ+Xz0WyKXnIF+qyu7xMTApjTlYkflGkrNnB8qJG5mZF8faKnjfQVqeHkXFqLpiWPPCGuXvJkHI0gXfv34VGRyMVtgr0Kj3JpmQc3uA+a+lswaKzcGHOhUToItjUsIlMS6YsVzzM6E8A4aUAgiDEAWcB/wPi+rOPgeLcyQn8sq0eh9tPslWPQgC/CEaNilHxZu7/Ng+X10+YXs3N8zJ58qd85udE8e+vcwFpXafHx6i4XeZTwzMkz9nRTHzlBqZFjGdl44bAYpPGhFf0cnbWWUTpo1hSuYS7l95Nq6sVtULNrZNu5YzMMzDspvqeDFC/DZwtEDmCqqIvsfQhk2BjvZZksxdfpxG1U/JFLVoLDc6GPToDOrWSRKueDRWtTE/b81TEcGJCchjfbarhrRVliCKkRhi47sgMVhQ3UdJgp2GnWhzHjxngdK+QMJh2LXz/957tGgO8cSKc8wZESU+OmZZMlIISn9jtwB+Xehz/2/A/qjqq0Cg1lLaV8lXxV4ywjuCRIx4hLSxtYO3fD8YnhlHUYOPpxQWBfrjz+By+2dQzA9ukU6FTKTnluaW8ePFkJiUPsJMTkQ0KZc/Rmkl/gtA9fxfyW/K57dfbKGsvQ0Dg0lGXMi9pHipBhVfsVro8LvU40k3pPLnuSZo6m1ApVNww/gbOzjwbs26Ap6RkDhh9jlYRBOEiQRBeBD4B5gPPAUcMlGH9YWZ6OI+fPZYkq55FebU8d8EERsWZUCoEnl5cEChL3Orw8PxvRTx4+mhC1AquOzKdaJOWUXEm/u/CiUzZ9UdpTYGLPoOs4zBWrOIfqadyRsbpWHVWpkRP4aYJN+H2ubFoTJS0FPOf5f+h1dUKgMfv4b+r/8vWxq0H98MYbmz6CFJmg6CgvKOCiL3ECwDkNmlIMXnwaqSRAYAwXVifBJyyo42sPARSDHfG7vLy5vKygNxtSaOdNaUtXDM3jczoUJYXNZEWYeC58ycweaBvPMD/s3fW4VGdaR++3/GJTNwT4iQEdy/FW6i3QG3rLrttt92tbm27X92FUqW69W6FlkJLKRR3CU6UhLjLTGbm/f54k4kCCYSQwNxcczFzbN4zJ+ec5zzye0g5D6Y9AZYIdeOf+jhs/RwKd8LyF8GuPGx9/Pswd8pc+gf0J8gcxKXJl6IRGjLKM7BLOy9vfJlxEeMA2Fm8k093ford2bodcHehymbnqw0Hmh2H7zYf4LIRvbhwSCQBngZGx/nz3OyB/N9POyistPGPLzdT1KK5UafjqFPHIGwQeIfC8OvAIxDqDu0ZqLXX8samN8goVx4NieT97e9jc9h4Y8ob9A/sT6A5kKv7Xs2goEHM2zqPolp1Xtmddl7a8BLbi9zXvp5ER57qXwT2AXOBJVLK9OMxoI5SVmPDoNVy4dAoxicGIZE4nZJ3rxzGuowSnC3qHwoqrNTanew6WMmDZ6Vw4dBIPA3aQ5de+cXC+fPAXkuMhz+j9gp0GlXbnlaWhlM6ya0uIK8mv5VimkSSW5VLrbUKk8MGJh+kEJRby/HQexxRZe+kx+mELf+F09RTZG5lDkOCD9+YqdwqKK3VEOLpwF7rhbFC/eYWg+WI5YWgkrz+2FPA36Yc+/C7C/sLqlpNW5NezPjEAF5YvIdJycFcOCSCMQldlDjpHQJjbgPvcNj5Hfz+H3Uj6n8RFOyA2nLK9QYMWgMDggbwyuRXOFh5kLuW3kVOVfOn6Gp7NTNjZ7LswDKWH1jOrbZb8TX5ds1+dJCMotYu9BX7irnvzD785/y+bD0Qyc/b87jxo/U46/tK7SuooqjKRoBXJysQ1pSCzggavWoAtvhhSJwGYQNgzy9QVQj9LkDqwiivysNstGBokqxZWFPIytyVrTa7r2wflyRfQqxPLFW2KmJ9Y9mcv7lV7gfQ6li66d50JEwQKIToC5wGPCGESAR2SSn/ctxGdxjyymv5cUsuH67K4LTEQAZE+vLO8v2A4JxBYZRU1XFaYlCzTmEAAZ4GtueUE+6r4lmH7HZXkQep38KaeWAJVzes6DF468x8vvtzZsbOJNgjmCVZS/A3+jOp1yQCzYEU1hQ224xBY+D91PkM1nkTavDjm/JdLMxcTN/Avlzb71r6BJzCyTYZy0FvBv846px2imuLjyg4tLvEQJTFjkaA3eiNV94OAHyNvuRVtZavbUlSqDev/r4Xm92JQXecyri6iJIqGyv2FuDr0dqoHBzlS2GVjWvHxfHC4t1cPqoLYtMNVBXCjh/AWqpKRif/C/Ysguz15J1+Nz+n/Y+1eRsZFzmO/+37HzX2Gv416l9oNa3zOGrsNfQJ6EP/wP7YnfZuXV3Qy791U55hMX4EW4wY9TqE0PD2srQW65gJ8OzEvKLyXNj2NWx4H0bfCtnrIDBReQd2/ti4XNwksqwlfLvjAxYc+IM+PnFcm3IFCcGDWHVwFcuyltEvoB+rD65utvkY7xiWZi1l/vb55FXnMSNuBpOiJhHhFcGBygPNlg31cKur9iQ6EiawoLoWRgMxgA/Qum9qF/Hl+mwe+yGVzOJqQn1M/P2LzaTmVpCaW86TP+3Cz8PA3vwKbpkQ7yrX8TBouWtqb7yMWqb3PYJ4yZbPVNyzaC+k/QEfnge5m+kb2I/rki8jwBzAe9vfI708nQ0FG3j4z4e5b8R9eOjUBUErtFzV9yre3/4+RdYS/le6g2ezfuKd1PfJrsxmYfpCblh0A9kVrTW9Txk2fABxE0EIcqty8TX6udTmDsWuIj0RXspV7DBa0NeUAuBr8mtXmMDTqCPS18zm7NJjHf0JZ/GOPLJKa9EKJXzTQKjFxKzhUXy+NptvNx3glUsGkRTahap9O76HH/4GB7fB5IdhyX9g329I4MvKvTy74UWGhg7lidVPsK1wG/tK9zFv8zzuHna3q+eAQHBx0sUszlzMs+uexeq0YtKZurWQjbdJ1+w4hPmYmDU0CpNePXMlh3rzj+lJNFQPWkw6nr5oYOd5BaSEjR/CL/dDr1Gw8jX1OXst9LuwyUBDqT3zSV7e+ibzdn1CdmU2iw78wQ1L72J7wWZu+/U2Pt/9OROiJhDsEexabVbiLDQaDXf+fidr89aSWZHJ3M1zWZi+kIdGPoS3Xv2NaYSG6/tfT9+Avp2zX266hI6ECZY3eb0qpTxhd7G88lreWa4s7P4RPqzeX9xqmdVpxUxJCeKX1Dxum5iA3SmRUvLi4j28feWww3cNq8iDla82n+a0Q85G/IZfy+zes7h88Q2AumhJJNlV2eRU5fCfcf8htTgVrdCyOGMxe0r3kF2ZzUMjH+KeZfc022SptZR9pfuI9I48th+kJ1JbDrt+hvPnAtRXEhw5X2BXiYEEXxVzdujNaBxWhMOm+hO0Q2sAlHdgxd5Chsd0rKyqO1FlrWNxah56nYZyfw/Kauq4c0oidqekvNbO/y3YwcTkID5dk0WglxEPQxfl+daUqnNHCNj6BYT2B2sFCEFeykw+ylhIpHdkq452fYP68ty653h87ONkVWRRVVfF0qylpBarJN/N+ZuRSM6MO7PbNClqyaasslbH4d8/phIT4EFymAVPo45rx8VyelIwJdU2ovzM9OrM7oUVB2HV6+q9dxgU7lbvd/4I0WPg9PsgpB+ED+JAbSELs5cCjdewcls5u+q7qEokL294mQsSLyDUM5ShIUOJ943n2z3fUteiQ+hnuz5jZtxM3p3+LpkVmfib/EnyS8Lb2H1ko90cmY6ECQ7bRFwI8YqU8vZjH9KRMWg1eBl1FFfZqLLaifRrXcIS5G0k3MeDPfmVzSRbdRqB55HkaLUGMFqgsoXbud5F6an3JN4Sx8VJF1PrqEUndFTZqzBrzRTUFDBvy7xmq3npvQjxDMFT50mVvXl816jtud3Kjont36j4ZX22cXZldrt6Euwv1TMhqj42KwR2ozf6mhI8vYKpc9RRY6/GrDt8D/U+J0HegFajwaDT4mHUoNMKVuwraiauFR/kyZTkEPIrrJj1XVnwI1U1QcVBFQIK7K1uQk47+rABJNtzOL3X6TicDm4ddCthnmFU2arwMfpw2+DbqLBVsL1wOz+l/9Rsq556T2xOG3rR/TwDxVVWth4ow6TXtHkc9NpG7QujXktKy6qlzkKrVzf7+Enq/QVvg9BAWRZ4BSutAUs4+ERisJUxJHgIYyPGUmOvwaQzkVWRhanJ9SjYI5ggjyDsTju5lbn4m/zbVFr1MniRV5VHhFcE02KmHZ99c3Pc6cyrxNhO3NZh8fM0cO+Zydzy8Qb25FcyZ3gUv2zPw+ZQUQujTsOk5CC+35zD4ChfNmaVuta9cUJco8LgofDwgykPw2eXN07zCoaIoQBYvEKZnXQRd//xT1dZ1NCQofgZ/aiwVRDiEdIsmfDC3hfy71X/5rnTn+WmxTe7pg8KGkSCX8Ix/ho9lE0fQ8JU18fsiixiLLGHXaXCJqi0CfxNjdEpu9GCvqoYm1cIfmYVKohuo4NaU5LDvHl1yd4erTdg0muZMyKK1ANlVNc5CPIyUtAkK33W0Ci255bRO9iLCL8u7EuwbwksuFu9H3cn/HQvlKYDEOATxUXn/B/3LX8AWa9sHucTx1lxZ7GrdBdf7/kagH8M/weLMhe5KgeMWiO9/XvT27c3Zn33ql2vstl5cfEePliZwWc3jGp1HG45PYH44C56QvYMhKFXwdfXQ+Qw8I+HzZ82zh96tUrqBCL8k5gQeRrPb3jBNXt4yHAu6T0bD50HDulgTtIcnlv/HE6pzrc4SxyPjn201fXtun7X8fDKh7E5bLwz/R13g6IeygnXCDhaJiUH8en1I1m5vwiB5PnZA9mWU4ZRpyXCz8y7y9NYk17CFaOjGZcYSF65leExfkzuE4xe245UiYQpcNWPqpueVxDEjFeJOECFrYK3tr7brD56UNAg5m6ei0M6uG/EfeRV51FcW0yEVwSLMxezu3Q3WeVZvHr6y2wq3EKsbyxDQ4a2S2TnpKM0S6k7TrjXNelA5QGGhQw/7Gr7S/WEe9tpqtjqMHq7ygv9jX7ktcMY8DDoiPL3YENmCWPie+7vPzLWHy+DltyyGh45J4W0wirKa+wEW4x8s/EAmUXVzBkRRW5ZLf6eXeCBqjgIP/9Tvdfo1KveEAAojRvP65vfcBkCAPvL9mMxWPh277euafO3z+e2Qbdhc9iQSOJ94gnzCuuWynZ78yv5YGUGOg2kF1bx3OwBbD1QTkmVjaHRfmQWV5FdUkWkXxckPlaXwNKnQDohfjIsfbL5/A3vq9wBSyg51Qd5c2tzD+bavLWU11XxwZkfkFGWwRtb3nAZAgD7y/eTUZ7BuQnnYnfaKbeWE+kdyarcVYwMHcn3+79nSeYStzHQQzmuxoAQwoRSKjTWf9eXUsqHj3W7dXYnWcU1WEx6JicHk5pTziu/7aWwspa/jIrhYGkt4xKDGBUfQGZRDV+sy8bLpCM5xKv5RbG6GErSQe+hxIWK9tWLc9jBK5SygHiy9DqkVo9G2tDkbULrtOPl2WgZB5mDODv+bOJ845idNBtR/8+gMbClYAtf7P7C9XWltjIGBw1hgtkPP5Nfs37tNfYassqzcEgHtY5a/E3+9PLuhThE85kezbavoddo5coErA4rZdYy/I5QMra/VE+oh6PZNLvRC321csv6GH0oaEcSIUBKmDcr9hb1aGOgoqYOIQTJoRbyy2vQa70pqbaxYl8RO3LLcUqlrFlldRx5Y8eC0wElGVCRCxMfALRgCVVhtgn/xGnyIdPTl1LfXhQv/wdnx53N6PDR+Bn9sEs7UspmN5286jxe3PAir056lVhLLDanjVpHLRW2CvzNKpRUY68hszwTKSW9LL3w0B/B23ecqLLaSQjy5G+TE8korub9lemcPySC2EBP0oqqkFJysKyWIC8Txs72QkkJxftUnoZ3KFTmQ/85EBCrSgrH3w07f4D8HdQF9yGz//nUiDoiK/OorTyIQWPgntH3oNfq0Qkdn+74lKq6KgYFD8LmsHFGzBlsL9rOkqwlrq8st5Xza+avTIueRqhnKPtK97EqdxUXJF4A0KqiwE3PoTONgbbuWlZgkpSyUgihB5YLIX6SUq462i8prbbx/p/pmA1avtl4gJ0HK9BpBJeM6EVGUZVqRpNexMerMrE7JX3CvPnr5ASeXrgLh4R16cUMi/FXynff3AC5m2H49VBbAhHD4M+XoCKXfTP+j/vzfie1eAc6oeOC3hdQYatgQOAActJ/5oKE8/kp42cuSLiAd7a9Q1VdFYHmQG4acBPPrn8Wp3Rybb9rsTlspJWrZMdE30RuXXIredV5mHVmHhr1EGfEnEFRbRFvbHqDRL9E3t/+frP502Omt7sjYo9h+9fNsptzqnIJMAUcsZJgX6mOEM8WxoDJgqlMXYB8jX7kVrVuDdsWKeE+LNiay93Tkzo4+O7BjtxybvtkA8Ni/An2NlJSZeO/a7OwOyV9wy3cPS2JZXsKyS6pPnJY7FiwVijjrroQlr+gPnuFwBlPQdoyagp38sOQC3hqywtEeUfx4KgHeWfrO3y//3t0Gh1zkubga/DloZEP8fjqx12bDTQHsr90PzuKd/DWlrewOW3EWmJ5esLT+Bp9eWXDK3y3/zsAJveazD3D7iHC+8itrzubuACl9vjA/7bxysWDOVhWy9M/7eKG0+JYl1HMmrQShIArRkVz66QEgjurlXSdFbZ/BT/+Xcmm+0TC6NtAb4I/noP87crYHno15QlT+czLwOt7/4t9/0ck+ybyTMq1PD72cf69+t8crDqIWWfm9sG3E+kVyTNrn+GL3V8gkQwLGca1/a7lnW3vIBAk+iYyKmwUb219C7vTTrxvPA+OepC3t74NqGPhpmfSmYXWL7WcIBUNwtj6+lerNsgdYUt2Gcv2FrJqfzE7D6oWoXan5MNVGYxJCMDqcDB/RYarW9iO3ArWZ5TwzIUD+HxdFrd/upHisgpY9pwyBEw+KtHJ4K2yoCtysUUO5e3yVFKLVQ27Xdr5fNfnJPgm8N7297AJ6OvXm9sG3cbrm16nqk4lBRbWFPLmljc5O+5sauw1vL75dc6OP5so7yj+Pfbf/HngT5dHocZew4N/Psi+0n2syFmBRPLxzo+bzX9g+QPsO0Jr3h5HeQ4U71dZ5vUcqDhAoMeRn9DTyvSEejZXoLMbLa4wgZ/Jj4NV7WtElBTiza6DFVTU1h154W5GldXOEz/uIKe0lgBPA0adlo9WZ7r+5rfnlLP1QBmXjezFgEhfauuOo2cgd4vyCCz5jzIEQHkEFt4HRi/2DLiAx7a/jdVhxc/kx4/7f2RPqUrotTvtfLzjYzQaDRLJ6RGnY9KaGB46nGv6XYNE8tqm11w9QdLK03hyzZPsKNrhMgQAfs38lcWZi4/fPh6GvAorD3y7lfIaO6U1dXy0OpNQHxNFVTbWpCmpbClh/soM1qZ1YhuXglT49ubG/ill2bD2bXBYlSEASltgzTxSo4fw8p7PXRLCO0v3sFun49l1z7rOlxp7Dc+sfYYDVQf4fPfnrlDOurx1VNgqGBo8lDuH3kllXSUf7/jYlc+xr3QfP+z/gRhLDI+NfYyhIUM7bx/ddClHNAaEEN8LIb471KthOSnl+4dYXyuE2ATkA4uklKvbWq697CuoZECkD2vTW5cTehr05JS0bqq4Jq2YHbnl7MmvJLesFntVEexZqGYG9oaDW8AzQJ1Q3mGUDpzDHwUbW22nzFpGmbUMi8HCx7s/RyCaaXQDFNQUYDGqbGGndBJgCuCKPldg1Br5Ys8XzZZ1Sic5VTksy15GkEdQK80BiTz5VLx2/aSSmzSNTqnsyiwCTIcvK3RKyK5o2zOgjAGJn9GX/Or2lRcadBoSQ7xYk9b676i7U1xlY/neQmIDPdlfWElZTWuDZsW+Ig6UVvPYD6nklnVicyIpwVrZqORVmgHOOhVaC+wNg/8Cgy4DoxeYfTlA4/nRN6Avq3Nbn/619lo2FWzi5kE3c8eQO9Br9Dyz9plWlTcA6/PWt9ngaFHGIhzO4xwOaYMDJdXU1jnx9dC5lCD7hvuwPqP1jX9NeifKYJe2UPzTGiB+KgT3VZVQDQgNWdbWf+MOnKSXpzebJpFklWe1WnZTwSZOizyNNza/wYGK1mGANblruGnATZyfcD5eBq+j2h03J572eAaeBZ47zOuwSCkdUspBQCQwQgjRr+l8IcQNQoh1Qoh1BQVHvpBH+nmwv6CK5DZEVMJ9TS5lwab0CbOwK085KHw99GhNPiokACpnILA32KpUjC15Jt652+nvl9xqO94Gb4xaI1aHlT7+KVgMFkSL6IjFYKG2STewgpoCfkr7iYLqgjZveA7pYHrMdCpsFW3ODzJ3kYQsHT8WR8WuBRDe/OkhqyKbgCMkUuZVafHUS8y65o4lp86IRKC1VeNl8MLmsFHdxk2kLfqEWVi2p/DIC54ADncsLGYdfcMtZJdWE+nngbepdbQvJczCH3sKMWg1BHaWqE3hHvj1UXhnKiz6l6pj9w4DnYcSFxp9m5LAzd2k2uMG9yWoSSlgenk6Sf6twzJGrZFE30Te3PImv2f/7vKUtVV2G+cTR7mtnL8N+Ru6JgblqLBRbSoYdgaHOxbBFhM6jaC02k6oRYUA9hdU0qeN61O/zmxR7NVE3a/3GXDeG5C3FZY9A1MfU6FPAOkkpA1dBi2aZoJCDYR4thZj6+3Xm2/3fkuNvYaANrRAkv2TT81E6JOMIxoDUsqlh3u194uklKXAEuCMFtPnSSmHSSmHBQUd+cY3MNIHHw89M/qH4ddEhnVyn2CCvY0M7OXLtJTGP2h/TwPT+4by5z51YXzmooEE+PvB5IfAI0D1WjdaVAZu3nZY+zbmjR/w1+DRzerex0eMJ7sim2v7XcvWwq2Y9WZW567mpoE3uQwCvUbPtf2u5cf9PyIQXJp8Kevz1nP70Nv5Pet3rup3VbML3CXJlzB/+3wyKzIprS3l6n5XN5t/26DbSPDtutLDjh6LDmO3QuZKCG/ef+BA5YEjGj3p5a29Ag04zD7oqwoRCALM/u2SJQZ1cV7eTY2Bwx0LH7OBR8/pCxK0QlBls3N6UuMy/p4GpqaEsDqtmP+c35/YwE7IZK8qhK9vUHkB+amw4mX44hrVu8M/TiWybf5UuarztsOG+fDro/R2arg67jwAlmUv46y4s5qJBk3qNYnqumoCzAH8lvUbEyInuIziLQVbOCf+HNeynnpPZvWexTPrnuHzXZ8zJ2kOAPE+8cyMm3ns+3gIDncs+oZb+OcZyWiEKjOc3CeYPfmV9AmzENHkwWR0fACj4zuxU2ZICpxWX7kxYA58fZ2S985eBz/cAaH91HUtaQYpJQc5O2KCa1Ufow8R6Pnn8H9i0jbmMPylz1+I9o5mWMgw17QIrwjifOJIK09jUtQkfIw+TI+e7ppvMVi4ffDtBHicPF1AT1WElO0L4df3Ivg/IAVw/QVJKQ/ZU1QIEQTUSSlLhRBm4BfgKSnlD20tP2zYMLlu3bojjiWrqIqfUw8SZjFTVlNHSbWN9RmlnNkvlNnDozhYXsOOnHLKa+308jNjtUsqrXaiAzyID/JC01CbVpKh5IZLMsE3Ej6+qPFLDJ4cGHUDaREDEBotOqMvoq6aCiTL89by9d6vcUonV/a5kpHhIymuKSLKOxIfJ2RVHsBk8kGr8yDMO5xI70gKawrJKs+izFbGloItGLQGlmYvZVvhNiwGC5/M+ISi2iIc0kGlrZIQjxDifOMw6Tq1RrzdpQntPRYdIm0ZLLgHZjzjmlRZV8XdS+/mr0P+iuYww/vvDi/SSnXMiG/tIg7Y8ytlUcOpiBzCD/t/YEz4GEaHjznicJxOyU0fr+fXuyYQbOnCWvwOHAc49LHIKKoivaganQaqbQ7qHE5sdqer/8bo+EBSwi3tK6U9Epmr4d02BGWuWahyBbLXqrK2llzwNtUaDXs9fcmzluLvGYxZ50mutQih0eKp88RitJBenk65rZxAk3rCLLWW4mfyI9QzlJLaEjbmb8QhHXyx+wvKrGUAPDfhOXyMPsT5xBHkcdTG6zEfi2prHdtyyskoqiba34O8ilpqbA5CfUyk5lTQP8JCn3Af/DuzBwEob2b5AfjzFdj4QfN5USNgwv1Qng0l6VQExLHPUUmVXzTRPnGU2sp5J/VjLkm5RAkKmf3JqcjBw+DB2Iix7C/dj91px1vvTXp5Op4GT0I9QimoKSDEI4T86nwq6iqItcSSHNDai3qUnISlUz2HjlQTvAc8DLwATASu5siehTBgvhBCW7/s54cyBDqCp1HH/BUZZLfIDzh3kCrVC7WYCbW0Q5zEL1q9frybdEsAu859GmnwIto7ioKqfMqtJcQLA0k/3odmwBxq05ayfMJfGRo6lAFBA6hz1uFv8qfUWkqSzpvk1IWw/AXih16pQg8mXxBG8Awj0BxIoDmQBWkLeGvrW4R4hDCr9yzGho/FqDXikA5SAlLYWbyTiroKDFYD1fbqzjYGTiz7lijVwSYcqMgm2CPosIYAQHqZjqBDeAbsRm8MVcp962v0bXdFgUYj6Bvuw/K9hVwwpOdJQkcHeBId4MmX67O4+4stzebFBHgwZ3ivzjEEwFUG2gyNDoRWvSzhcNo9oDPB5k9UmS5A8V6qyzIpTplOibTigZNaRw1F1hI8dB7EWGKoddSyKX8T/QL7kV2RjZ/Jj8yKTBJ8Ewgyq1waicSsM+Nv9KfMWoZGaAg0B1JqLWV17moSfBPo7d8bjej65lMeRj0jYgPIKKpm9rzmhVIxAR5cPGIsvh7HoSLIVgW1lWCyKIEnnVF5OBGQt53ygu3s1DrJ9fIkzFZE8u4lWCY+CAG9KS9KJbcml00Fm7A6rAghkEJSaatkZ9FOKm2Varqnkio2aU0U1RZRXFuMUWtkQNAALEYL+0r38XPaz2iEhmT/ZHpZenX+frrpEjpiDJillL8KIYSUMgN4RAixHvjXoVaQUm4BDt+T9ijw9zJyz/Qk/vbfTY3TPA30jzi6mNyuwbO47o+7KbWWAsqNdnXfq3lx88vohI43JtzByJ2L+WHoLHYXbWV93np2lyjdb53Qcdewu3ho73z+PfCvJNWWQMZK5S4FdcG87AslEQqk+KcQbA7myr5X8sL6F1w63wvSFnDHkDu47bfbXOOaETuD+0beh6/R96j2q9uxfwn0u6DZpOzKI+cLAGSU6+kXaGtznt1kwVCp9AX8zQEc6EDzp5QwC0t3F/RIY6CBQVG++HroKa1uTCS8e1oSfp35JBqQCH3PVzLSDcx4FnZ8B1mr1QuU/O2UR5WXIDiFIpOFR6x1LF3xECatidsH387z6593CXb18e/Dv8f8G5POxP3L73dtek7SHHyMPty77F7WHFyjNo3gjqF38NaWt7hhwA08sfqJxvNQo2PelHkMDzu8cNXx5FDH4bgYAlVFysuW+i1c8R18c6Oq6gAwelN7yad8mPsHc3f817XKTb0v5jrfXhiBBJ8EZsbN5Om1TzOp1yR2FO1g2YFlrmVvHngzIR4h3LToJhL9Eukf1J/Pd33umn9d/+uYFj2NaxZeQ2WdyscKMgcxb9q8Lg1tuuk8OmJGW4UQGmCPEOI2IcT5wAlLHZ3SJ4T51wznyjHR3HtmMp9eP5L44KMbzk+5K12GAKiqgfTydCK9IrFLO89mLmD/sCt4evvb+Jv8XRcgUGWHX+/5mmifGLbW5EBAEuRta9y40w4/P6CEQYAYnxjemvoWC9MWNmv4sbd0L6lFqc1q7RekLWBvyd6j2qduh7VS9bIPau5SzCzPJMh0eGPAKeFApZZgj7Y9A3VmXwyVyjMQYApot2cAYECkyhtob7isO5IQ7M1/rx/FvWckc9WYGD68dgQTk1snhx0TJm+Y9gSc/6aStb3gLVW6ZvRqNARAPZmumQcXvAlDrmC3TxBL69vgTu41mS93f9lMuXNH8Q52le5i/vb5zb7us12fkVGe4TIEQGW7f7LjE54a/xSB5sDm56HTzosbXqTSVsmJIjGk8ThcPSaGj64dwaQ+nXwcGsjfrgwBD3/Yu7jREACwVpBRkcmbuz9rtsq8PZ+TXqY0TwpqC3hj0xuAMsiaGgIAH6Z+SIm1BLu0MyFqAl/u/rLZ/He2vsOu4l0uQwBUsvTvmb933j666VI64hn4G+AB/BV4HJgEXHk8BtUePI06JvQOZkLvDp5slflKNtXsB75RAK06qEV5R5Hsl0ySXxJr89ayNnctlQYTVocVm6P102luVS4jw0ZSXFsCttaljZRlQF0VmH0BsBgtZFe2fnqtqKvAoDVQY2/cRrmtvGP7113JWg0BvZUbuenkimxGh4867Kp5VVq89BKjru0btt3si6GqEJD4m/worCnEIR1HFDECCLGYMOg07DxYQZ+w7tkNrz0kh1lIPt7j94mAgRerF8CqN6GtUrLyA1C4Dxy1lBsbPS4B5gDyqvOY0msK8b7x2J12FmYspNxW3sxAaKDCVtFqWn51PnG+cfye9XuredmV2dQ4avA6cc8ox+842KpV5ZNGp5I2a0qVuNPIm5WR3YIKa0Uz2WdQpcwV1jKoLaemroaKugadFjsBpgBmxs3EpDORX53Pgv0LXFoCLRUiQRlmJdbW5ZMtyxXd9Bza7RmQUq6tFxAqB/4qpbzgWJQETwjZ6+CdafDmeJh3mpLqdDo4K6axwGFYyDAm9ZrEixte5Km1T1FUU8SDIx+g167FJPsm4qlvnZk9MWoiq3NX09sSo6SNWzLwUmhSshNgDuD8xPNbLRbpFdnMEDBqjURboo9tn7sL6cshuLlXwIkktyrniGVJmeW6Q3oFAJw6E1Jo0For0Gv0eBss7ZYlBhgQ4cMfu49TKeXJSl4qZK8Ge60KDTQl6UxVcqj3IEbvi06oZ47Vuat5YNQDHKw+yJtb3uSTnZ8wLXoaCT4JhHqGNtuEn9GPSO/IVjkA02KmEWQOok9A6z4F5yWcd0S9ih5JSTr87xZ4Y7R6/fYY+MXAiBtUVUfksFarRHpHtqrQCTIHEbl3CXwyh1CnZGhwfeM1o4XLUy7n273fMm/LPDblb+L+kffjUd/9s6H5WlMCTAHE+8a3+t5JvSZ1zj676XLabQwIIYYJIbYCW4CtQojNQoieIzdVkQ9fXqvaeILqS/D5lVC4h1EOLXf2vgQPnQdjw8cyf/t8ah1KK2BzwWaWZP+OqaaEf4dNJrsik78N+RtB5iC0Qsu06GlEekdycdLF+EkdxJ4Gsz9QF0ONVomwjLkNtI1OGCEEs3rP4uKki9FpdASYAnhy/JOMDB3J+IjxAMT6xPL65NeJ8zlksUbPIuNPCE5pNqmwugCjzoRZd/hkz8xy/WGNAYA6Dz8MFcoACDQHcqCy/WJN/SJ8WLKr/caDG2Dnj7DtK9j0CZzzipLDFRpImgEhfdUNq66WhIoiXjjtSSK9I/HQe/Brxq9sK1RhtBp7De9ue5esyizuH3k//QOVKmUf/z48c9oz9Avsx8sTXybMMwyN0HBmzJncPvh2jDojfQP68tRpTxFgCkAndMzuPZvZvWefkATC487WLxtzNZx2ZQBU5sOSJ8BWqQSHxt2pEpb1Zhh2DaGVxbyccj0DApSsywD/FF5OvIzQte9D5grse3/ljNgzGBU2CrvDzisbX3F5IdPL05m/fT4+Bh/Ojjub7/Z9x/X9r2dAoEr+7R/Yn1cmv0LfgL7cMvAWTFoT3npv7h1xr1uBsAfTkTDBu8AtUsplAEKIcagKgwGHXau7UJHTrIMaoE6smhL81r7D1YW7mTFwFgs1rX+SP7L/IHfwPcSvmMcZk+5mc0U6N/S/AaPWSKBHIMW1xRysOshmey0D46eqGuBeo6Cutl6UpXUCUbhXOP8Y8Q+u6nsVBq3BVRr13ITnlIqhwYLvERr39BjsVqXyOO7vzSZnVWS1KXzSkv1lOoI87Iddxm72w1CZT01gAv4mPw5UHmj3halvuA+v/76PKqsdT2OPbeTZNTgc6rxpUPAs2qtc1vGTwDMI0pbC0qfVvJixaD2DMVfnMDR4KGMjxvLg8gdbbTKnMof/W/1/PDDiAZ4Y+wTeBm+XPPWEqAn0D+xPtb2aIHMQRp3S4TDpTMyIncGIkBFYHVaCPYNbCYCdFFgrVS8PjU55AOxWyNmoBJ+kU/X4SFuuNAYGXaIMg72LcTjt9Isex1y7D2Vjn8Vn/Yd4f3+PaioFFNgreWL1a4wIHYFJZ2oVBkgrTyPMK4x/jfoXtw66FYPOwIy4GZRZy/Ax+OBt9MbhdHDjwBs5L+E8NELTpmCRm55DR658jgZDAEBKuVwIcfgrdHfC7Kf6ENSWNU7T6FTJVHBfxN7FhC59Dv9L3m61aqJvImVmX+7oFYtj/zcMDxvJl3u+ZFfJLgD6BfYj0BzIufHnNq7kdeQTQ6/RN2uukluZy5KsJSxIW8Dg4MGcG38uCX4nQWZu7mbwiQJD8xBKZkVmu5TLMst19PFvu5KggTqTL8ZylUQVZA4isyLzsMs3xWzQkhjixcp9RUxJcV/Q2iQvFTZ9qkIDw66B8CEq92bUzSo0tqFFnbvRAsXpYPDC1yuE7/d/T2VdJfG+8ewobh7jNmqNXJ5yOZmVmXy972vGho9lRtwMV4jM3+yPP/60RaBHIBllGby79V2WHVjGhMgJTI+ZfvKUuOnN0PdC0GhUaa7JB6Y/oUqi53wE6X+CwRO2fQGr3qAk5RxWDJvDl0UbiS5ey0VRQ+hnrYYd3zfbrEXoCTAFsObgGkaEjmj1tf4mfywGCya9iQh94zXK2+DN9sLtfLH+C9LL0rmw94WMCx+Hn9nvuP8Ubo4vHTEGlgoh3gQ+RTUbmgP8LoQYAiCl3HAcxtd5+EXD2S/DV9e4rGPOnwdfXQ8DZqmknJI0+lWWMSZsNCtyVwJK9ezGATdw1ZLbXdbzHweWc/ewu9m9bjceeg9mxs5kZc5Kl5vzaLDarczdPJev934NqPDEz2k/M//M+YR7hR9h7W5OxspWVQQA6WUZxPnEHnbVhp4EwYfQGGigzsMP74OqQUuQRxCrD3asBUb/CB8W78hzGwNtUZoJH12ovGugkkHPnwf+sfDLgxA+CIZf11hOKzQw9m/w26NgtxJ73a/cPexunl/3PH8f9ncyyjOotivxqNHho9Fr9GzI28Cmgk2A+tv/Pft33pj8hqtl8aEoqS3h/uX3s6Vwi2vdFTkreGniS64eIT0ajVb9zl9e3ThtzyIl9vT5X5RBNvYOCO4D1cUsiO7Pk6nqOKwDFuj+4KPJb9I7aSbs+lGtb/AiJHosj0YN4o7f72DNwTWcG38u/9v3P0CVS9897O42xYT2lOzhmoXXuI7f+vz1/GPYP/hL378cz1/BTRfQEWNgYP3/D7eYPhhlHHT/zJHkmXDjMpWQ4xUCtaVQsh+WPauedjwCiEXPwzHnsTNmBlUOKwmmQL5I+6mVG21D/gbmTZ2HEAKd0DEzbiZ+pqO3jrMrs/l237fNph2sPsi+0n0ngTGwXD1JtiCrMpNRR6gkyK/W4tFGT4KW1Hn4Y6jMQ1UU+FNSW4rVUYtR2z7RpsFRfjy9cCdSSoQ4Cd3Nx0L+jkZDoIHFj8KQy1TIIHudMgAmPaj+9whQce0K1RHPcHArF/W+iMHBgympLeH1ya+RW3kAL4M3Xnovquw1vLDhhWabTy1KJb08/YjGQHpZussQaGBd3jrSy9MZENQzIpiHpa4WVr/ZfJp0qjBB/e/LipdgyFUUJExk3pbnmy1aY69hR2Umvc95BQpuUWGHwAQISGCc08HnZ31OVkUW/iZ/psdMp6C6gCjvKPoHtP1gs6t4l8sQaGDe1nlMj53erpCfm+5Lu40BKeXE4zmQzkZKyZ7SPaSXpeOh8yDJHExQwR7VbS2kr7K29/+uFnY6YM1bWMMGsHviPWQX7yBA6Bi6fwU+xenYU8Y22/Z5CefRx78P5RU5xFiiSQwejNBosDls7C7ZTVaF6sKX5JeEj6l9QkgaoWkz5tnjb0xOJ2StUZ3smlBhq6CmrgYf4+F/n/QyHaFH8AoAOPQeIJ1oayvAZCHIHEh2ZTbxPu0Ls4T7qoYz23PK6XeU4lUnBXW1qoa9OB28giGkH22qxDptSnmwgaw16hUxBHxj2D/4Yvbq9ZiEht5eAYQ66oi1OqhwOtlfug9/cyBJ/n0I8wpje+H2NodSZi0jryrvsLHoQ50fJ00ioRDN1R+DkqH/LJU7MPEByNmgOoGufw+NTtdmOa2mrhqKdkPYQDB6k16Wzt6MxXjoPVyN15zSiVZoMevNaISGDYUbqK6rJsE3gRifmMZttfG7aoX25MzXOMVotzEghAgB/gOESynPFEKkAKOllO8ct9EdA+vz13PjLze6eqEP8U/hSX0vwlbNU16Bv3yrQgPBKarxiiWCn4fO4sGVD7i2cXGvafzNUcd5MWfyv6xfcUons3rPYkfxDr7d+y2g4v7zJjzPsF6nszB9YTMVtdm9Z3PH0DvwNrTuYNaSSK9ILkm+hI92fOSaFmuJ7flqXoW7Va5Ai0YmGeWZhHiGtkOGWE+wuR2pKUJg8wzCWJFDtclCsEcI6WXp7TYGhBAMifZj4baDp64xIKVKVvv25sZpI26AUbeAf7xqRtRA/1mqXa7OpMoLGxh6DVtFHdelvul6gux9sDev9b+dP2oO8PjaJ12Lnh13NncMuYNoSzSToibxW9ZvrnkDgwbyY9qPvLX1LZ6d8OwhvWOxPrGMChvFqtzGKufTI08/eUpydUYVBkhfpq5b/S+CJf9ubCHd5xxInAZ7fiFA68GtiXN4ZMurrtUtBgspxQfgs5tg8iPs6DOd6xZdj16j57I+l/HKxlcYFzEOo9bI4szFgFJ6vG3wbXyy4xPqnHW8Ne0tUgJUJVCyfzI+Rh9XfwiAWwfdeiy9Idx0EzoSJngfVT3QcLfcDXwGdDtjoNJWyfPrnncZAgAbilPZnjKWMIDKPFUW1eds1f4z7nSyghP5z47m7rj/Zv7CzAkvMiB4KO9Nf48FaQuI94nni91fuJapc9bxzKZXeca7F/9Z/Z9m63+++3POij+LwcFHVmTWa/Vc0+8a+gb0ZUnWEvoF9mNi1MRW9dc9jswV6qbRgvTyNELM7agkKG2fZwCgztMfU1kO1UHJhHgGs78sjckdGOqwaH8+XJXO36e3brN7SlCSriRum7JmnroBXfqZaj99YIMKt9XVQMFO9XSan6p08iOHYTX68GbOT81cybtLdpOl1/PCylebbfr7/d9zRuwZnBZ5GveOuJexEWP5I/sPYnxikFLyQapKStxSsOWQxoCP0YdHRj/CsgPLWHNwDaPCRjEuYly7DPAeQ8w4JTlckq5aRzdVy9zxHcx8XlVzOOqYtn0h/n1v4ufyvUQZ/Ziq8yP+V2WAObd/xX9FMeW2ci7vcznzU+cjkQwMGsirmxqPjUQyf/t8zks4jw9SP+CTHZ/wyJhH0Gl0xPnG8c60d1iUsYjM8kzOiD2jWZdDNz2XjhgDgVLKz4UQ9wFIKe1CiPZdpbuYqroq0svSW00vdFiVy81Rp8pzEqfC8ufB7Efl5HtbxcIASrVadB6+DPEYwpCQIXyxbX6rZdIrsqiyVzWT5nStX1va7nEHeQRxVvxZnBV/VrvX6fakLWszeXBf2X6ivI7cDyCtTM+gYGu7vqrOHIipVCk7hnmGs6ngpw4NNTHEi9LqOvbmV5JwlNLWPRpruapbb0l5LpgDlNqdzgDVJZD2u2pVvOZNJYCjN8PO76kZfSt7K9JabaLKYW3z/CipVSp2YV5hTI6azC/pv/D17q9d6ngAB6sOHnbYEd4RXJx8MRcnX9yh3e0x6E0QMVTlY7R1PRFa+P1JqC3FG5i4/3cm9jkXtNWw5TnXYjbvMHbWSzh76j1dT/dNH5oaKLeVu/Q/dhTvwOawoasvu07yTyLJ/xQ1mE9iOhJYqxJCBKCSBRFCjALKDr/KiSHAHMCMmOmtpidozMoQAOXmNPuri1hNCWHZm0iwNM9sN2gMRHlHNZsW793a/TgjaiKhnuH09uvdbLpeoyfKEtVq+VMGKSFjheqt3oL0snTCjpAYaXPAwaojVxK4lvcKxFSqSgoDzYGU1pZQ1cYN6FBohGBUXADfbTrQ7nVOKiyRENjCcNMZlQfg9RGw8AE4sB4+nqWy2y31x68kXSUZSomPJZqz485utelAjalVyEuv0Tdz5/uZ/UjyT2pmCABtqg2eUuSlwo93wbp363M4mqDVK5nopkaC06GMhwPNWy2bStI4J24GAPvL9pPir1z/OqFz3egbSPJLckkLnxt/Lh5tKau6OanoiDFwF/AdEC+E+BP4ALj9uIzqGNFJyRU+KUwPG4dAYDFYeHzwHfTdvkBd3Mbfrdxq1iqY/C/w7YXv1i/5v4RLXIpd4Z7hvDr51VYKgCnBg/j38PuwGCwIBFMjTuOqPn/BzyOA/4z7DwMDVdFFqGcor0x65eRREDwaStJUtrl385t+ibUUm9N2xG6MmeU6As0O9O38K60z+6G1VqGpq0YrNER4RbCntGONnkbHB/D1xgM9unHRUeMZABe+DRH1bl/faJj8MKx/T91gKg+qMMKBtcrQS/8Dxtyu+hNotDDkKkTyGZyXcB7nJ5yPRmjw1Hty34j7iNGY+NfgO+gboEJGoZ6hPD3+afoFNt7cNELDxckXMzNuJhqhwVvvzcOjHz6mkt2Tgu3fwJbPVEhg0CUqSRPAEgGTH6n//2Gl96DRwcibVIfJ816HoHpDKqgPnPsak2Omc2nypfyR9Qcz42YyKGgQX+/9mruH3u2SHB4YNJDzEs5jaeZSLk26lGkx007MfrvpUkR7L3pCiFnAQiAKuBAYCTzUmfoCw4YNk+vWrTvygkeiPAdeH0VtaH/y4sZjcDoIK89XUqlBvcG3l7p4bfkMfrgT+l4A3iFQtI8KoGj0TXgHJBJgPrTO+cGSvdTarYRaojA1qWeusFVQVFOEl97LpaLWjWh3ym+nHIv170Pq/2DcXc0mr8tbxy8Zi7ggoXV/hqYsTDPzR5aZ2cntf7oP3v49BSlnUh2UzIqcFRh1Ji5NvrTd60spuffrrTxz0QBGxh03nfsOpV532nnRXmrKVF5N2lJY/Ih68tToIekM+OPZ5sv69oIpj4DODAlTQacy320OG7mVuei1+sZ4f1UhBbWl5DmqsJj8DikMZLVbOVh9EIPGQJhX2PHbT0X3PhZVRfDBOarCI6Sf8mwGxEPv6VCcBmvfUTf95LOUHoR0KIGvhgqE6mKoKgTPQNXhEJXnlFuZi1Zo8TH4UFhbiFd9w6lKWyUBpgDK68pxOB2EeYWh1+gPNbrOxl2ScALpSM7AQ1LKL4QQfsBE4FngDZRRcMLYV1DJz9sOsi6jhOkpIUxICiLM5A2BSZgyVhDda7RSHSzLVJoCYf2VIQDgEagSnzZ+6Nqet8EL77F3qhjpocjZTOi2L6F4Pwy4GGLHuzoSehu8T67kpWNh32+t3ZrA7pI9hHse+SK/p0Tf7uTBBqxewZiLM6kOSibaEsOv9RnS7UUIwYTeQXy0OuN4GgOdht3hZGNWKV+uz8Za52D28CiG9vLDqD9yx8ZDYvZRr4w/YeJ9kL0eHFZVURAzXmW2N1CaCUX7lHcgfal6As1ciWHrl0SH9IV+F0CDMeAZSJBnIEfKOzfqen6DrhqbnTXpJXy5LguLWc+FQyIZFOWLRtPB+53BCxImw4A5kLlShTXDB6teK2vfUXkeXiHqXNvymUrs7D+78Zrk4e8yAhrQa/TNDDEvY2N+TIMiqLfRfQ071eiQHHH9/zOBt6SUPwoh/n0cxtRuDpbXcOMH69lboJ4cl+zM5+LhUTx6Tl+MUx9T/b63f6P000GdMHnbYcazKinHPxaGXgPr323c6OR/qVa7hyJ/B8w/S52EoDofnvUSDLvquOxjj8XpgLQ/VKZzC3YV72Rc5PgjbmJXsYFJvVondR4Om3cInoV7KALCvEIpt1VQUFPQqoPb4TgtMYg7P99EQYWVIG9jh76/q9mUVcrF81bhcCoP37ebcvjw2hGMT+yEUi//OPjo/EbFzp0/wkXvQ+YqcNbn3vQ5G6qLlADOmnkwYDYsfUrN2/0TbJgP1/wCAadWuOzPfUVcN7/Rg/D5uiy+uGk0g6I6KEymNyrBri+adIvf8R2cN1ddg4Zdq4SEPpmtQnIN82fNh77nHfuOuDll6EjOwIF6OeI5wAIhhLGD63c6e/IqXYZAA5+vyyKjuBqiR0PshEZDoIFNHykPAainGoMHXDAPznhaXej2L4HS/Yf+0tzNjYZAA0ufVJa6m0ZyNqouap7Nb0pV9mryqvMJO4JnoM4JGeV6Irw66BmwhGEqzQKnAw0aEn0TWXdwbYe24WXSMSY+gPdXpHdovRPBd5tzXIZAA/NXpON0dkLOw55fGg2BBjZ+BDetgPPfhAvfUaW51gpY944yDFa80nz5qgLI23rsY+lB1NoczP19X7NpdQ7J77uOok223Qpr3mo+zVGnrkPXLoYpD6uSzwZDoIG1bysvgRs37aQjnoHZwBnAs1LKUiFEGHDPEdY5rrSV7iCbTtcdQoq2Yb6U0KL2GSFg0kOuj5XVBewtTKWwppAI70jiNTpa9SCUThCnYMLZ4dj1s8pobsHOoh1Eekega0MprSlppXoCzQ6MR5AhbolTZ8RusmAqzabWP5o+AX1YkrWEM2PPpCMhyRn9w3j4u+3cMD4OH48ui5l2mLZyfhxOSaf8NbZ5gjnUU35wb5U/UJoJvlEw4Z8QmARtfXN9BU+ds459JfvIrswmwBxAom+iK1Z9UiHA2cZv19a0IyIl4Gw9XauHqOH1y7QxXzqptVWxN28DB6sOEuoZSkJgP0ztVER1c+rR7id7KWW1lPJrKeWe+s+5Uspfjt/QjkxiiBexgZ7Npl00JJLoAFUfS3Ay+LdQoBtwsQoPgKp/D2wREug/WykTAtXVJby35R3+suQ27lz1CHMWXc8ijRUCEpuvc9o/2tWl8JRi53cQ1bob2uaCLUR7xxxx9dQiA1HedUf11bU+4XgUqnrqKO9IpHSypWDLEdZqTojFxLBoP17/vWPVCF3N2QPDaRmGvmpsLNqOxqbbIuVsVdvelNG3grb+GSL2NFW+9vuTKjSw4mUYdWvz5T38oTIfSrNZkrmEOT/O4c7f7+SKn65g7ua5VLala9DDMem13Dghvtk0nUZwetJRaPfrTUoFsikaLcSd3vg5aUZjHlQ9dWP+xld7v+aSX2/izlWPcMmvN/HVrs+os7m9BW7apkc3bw/zMfPWFcP4YUsOq/YXcdaAcCYlB2PS1++WJRwu+VTlDaQvU+U2vaerJBwASxhc/DFsa5h/HiROd7Xa3VecyrxdH7u+TyJ5fPMrDDjjcaK2fg0VeTDwEuUqddNI4R6VBd1CbMiJZHPBJi5OOrI4zJYCA9GWo+uQbbVE4JW3neLeUxEIxoSP5dNd/yXZPxmDtv05ABcMieT+b7Zy8YherYzO7sLgXn58ev0oPlmTidXu5LKRvRgec/jmPu0mYjhc9aMqLbRVq2Ze0WMa5xfsVjk4DeRuguHXw5THYN9ildUemAi/Pc6B0BQeXfVos4Zf81PnMyV6CoOCB3XOeLsR4xICee+q4Xy8OgMfs55LR/ZiUKRvxzfkdKjfeepjKgdHb4Zeo6CsSeOomPFwyWcqBGqrgcGXkeETyLOr7222qWc3v8GI4GEkhrVuGubGTY82BgASgr24Y8phEv6CesPp/wT+2fb8wEPPL64pbDWtqq6K8rwtSpbVw78+Y9fdy7sZWz5XEqotnip3F+/CQ+95xO6OUsK2AgOnRR7dU0ytTwQBuxejqavGqfcg0S+BfaV7eWH9i1zT72qC2tldzd/TwDkDw/nHl5v57IbRHc8E7wL0Wg0j4wKOT+WDVqdu/k0NgKYU7Gw9LX877P1VnRcZf8ImZUxXWEsot5W3Wry4trgTB9x98DTqmJgczMTkY+zkV1cDO/6nEpfDB0GFFXZ8rxKfB9Ub1XqjUlNNnOparSRtEXbZ3Ji2SzultUXHNh43Jy0nSWuv40Okd69WNbaRnuGEluWqxMScjUqYxU0jToe6AcS1bnK57MBykv1bSxO3JK1Mh0kn8TO1EQttB1Kjw+obiVfeDkA1XpkWM51gjyAeW/UY9y+/ny93f0G57cgCmmf0DaXSamfu0n1HXPaUI6YNI8EnCpJnQNZqpUxYT4hnGLE+zRU+dUJHpPeRJalPaYxe0G+Wygs4sKHRExN/+CayYd6RWAyWZtMsBgvhLRRV3bhpwG0MHIbYoP68OPYJAkzqqSvWEsPTSVcSsOm/SsRj9ocQ5NbobsaeRWD0VsIoTaiwVbAxfyP9AlrrDrRkTa6JRL+jyxdooNovBu8DG12ftULDmPAx3DLoFqZFTyWvOp+Hlj/Ennqt9kOh0QhuOT2Bt5btZ9meo8gGP5mJGqUaFemMKvG27wXKEBhyJfQ5Vy2jM8Hkh/ELSOLJ8U+6JIn9Tf48f/rzxPvEH+YL3ADQ/0KlMyAEaA1w2j2qWuowRAb24aWxT7ganYV5hvHS2CeIaCk37cZNPe1WIOwKulzdq53kleylrLaUIM9Q/LRmJctq8lXKbD2L469A+O4Z0GtMqyeXL3Z/QU5lLtPbIW1666JAJvaqOSaDQGO3Eb7hI/ZNeQCnoe14f1pZGgvSfuKe4XcT5d22Gl4DO3LLefm3PXxy3ShSwi2HXbYddG/Vu47gdCgPgLMOfKLBUJ+PY6tWlQY6A/jGgEY9d5TWlpJfk4/FYOkuHTl7xrGoq4XSDJUo6BvTmMR5BApL0ymuKcTfHEigb8xxHWIn0P3icJ2EEGIBcKmUsvREj+VQHFfPgBAiSgixRAiRKoTYLoT42/H8vjYp3q/qdL+5CbZ8oZL+OkiIXwK9w4bhZ4lU+u0hfXuiIXD8Sf9TXbBimwsK5VfnszR7KaPDD/80A5BTqSW/Wkuc77F5Bpw6AzV+0fhkrT/kMrE+sZweNYFXN75GrePw+Ql9wixcOTqGK95dzd78isMue1JStBdWzYVvboZtX6sKAVA3p4B4lSzaYAiASsINTlbCRZrGy4yvyZfefr27iyHQM6izqqqNNW/Bxo/h4Oa2yz7bINA3ht5hw3qCIXBSI6Wc0Z0NATj+YQI78HcpZQowCrhVCJFynL+zkYo8+OIqWHA3bP4Uvr5O1Ubb29cS100HcDpg4X2quqJJB7Q6p525m+cyMnQUlnbINP+wz5MhwVa0nfCMUBmagm/asrbrsOvpG9CXcK8wPkn99IjbGxUXwOxhUVw8bxW7Dp5CBkF5Dnx6Kfz8T9j8iepYuOoNcBxdtYebDpKxHN6fCWvfUi3X3ztT5Su56VSEEJ5CiB+FEJuFENuEEHOEEOlCiKeFEFuFEGuEEAn1ywYJIb4SQqytf42tn+4lhHivfvktQogL66enCyEC699fXr+tTUKIN4UQ2vrX+/Xfu1UIcWdX7/9xNQbqtQg21L+vAHYAXfdIXbBDKXU1Zd3bylvgpnNZ85YyCGInuCY5pIN5W+Zh1JkYFjrsiJsotwp+SfNgVERtpwzJ6hWK1Jnwztl82OUmRk1ke/F2NhVsOuI2xycGccmIXlw8byXrM0o6ZZzdnrztULir+bSVrzRLEHRznLBb4c+XW0/bvfDEjOfk5gwgR0o5UErZD/i5fnqZlLI/8CrwYv20l4AXpJTDUY373q6f/lDD8lLKAcBvTb9ACNEHpeI7Vko5CCXzfxkwCIiQUvar/673js8uHpouSyAUQsQAg4HVLabfIIRYJ4RYV1DQyQlaLaVUQT0lHuZJ8VTmqI/Fwa1Kknn07a5ywjqnnTc2z6W4tpiZsTMRgM2hXm0hJczd5MOgYCv+R1lF0AohKIsaSlDqjwiH7ZCLGbVGzow9k/e3vUdJO0rdxsQHcsNpcVzz/loWbM3tnLG24LieFx2lpdQtqHNLdkwquqdyQo+FlG3LCrulho8HW4GpQoinhBDjpZQN5UafNvm/IdY5BXhVCLEJ+A6wCCG86qe/1rBBKWXLJ4bJwFBgbf26k4E4YD8QJ4R4RQhxBtC6Dvc40yXGQP2P9BVwh5Sy2U5KKedJKYdJKYcFBXVCc5WmBPYGnxaJYSnngV9M537PScJRHYuSDPh4lhKbqc+jKLOV8fTap6mqq2Jo8Cze2uzHFT8Gc/43YZz/TRiXfR/C06t9+SPLREG1hvxqLa9u8GFXsZ5psVWduk+1PpHYvIII3vbdYZeL8opkUNAgXtn4KjbHkcNIg6L8+OcZyTzy3Xae+HEHNnvnGpjH9bzoKMF9wKtFvfzAS0+Z8+iEHgu9Ccbc1nya0EDSmV07jlMAKeVuYAjKKPi3EOJfDbOaLlb/vwYYJaUcVP+KkFK2R05TAPObrJckpXyk3mgYCPwO3ESjp6HLOO7VBEIIPfADsFBK2bqFXROOS6Zu/g7VojhjhVIgTDkP/E5ZbYDOrSbIWKm6qfW9AJJnIpFsyNvIB6kf4qmbyM6iUewt0TM8zEr/QCsh9e2Ii2o07CkxsKdET3aFDiGgj7+N6bHVeOg7/+9RY7cSvO1byiOHUJQ8nUPZwBLJz2k/U+Oo4bZBt+FtOHLVQHlNHW8v309JdR0Pn53CuIRAhDjiz9wzMtibkrcN1n8A2Wuh/yzVlMj3pKhZ7/7HwloB+5fC6jfAaIFRN0Ov0ao/wcnFCa0mEEKEA8VSylohxFnAdSj3/Vwp5ZNCiMuBOVLKs4UQnwAbpZTP1K87SEq5SQjxJGCSUt5RP91PSlkihEgHhgHBwP9QYYJ8IYQ/4A1UATYpZbkQoh/wUX0Yocs4rsaAUFfF+agf+I4jLX/cTjQpVZxNf4jGRacOx2YMOOpUMlnORtjyOTJrNdZh15FmiWJdXha/ZaZxoCKUoupBGHVaRoXXMCTEir4bqFlobdUE7l6EU6OlNG481f4x2E2+rTTdnThZfuBPthZsZXKvSQwMGkioVygmrbntDaOaBa1JK+arDdmY9FrOGhDGsBh/YgI9CfQyYNS1asrU/W9AbXFynkc951g4bICm3WWFPZATbQxMB55BdYaqA24GvgQ+A84ErMAlUsq99cmArwF9UEq+f0gpb6r3gr+GCgU4gEellF83GANSykIhxBzgPtRTSR1wK1CDyhNouFreJ6X8qQt228XxNgbGActQbpcGP+r9UsoFh1i+AMg4TsMJBFrrC3cfumJ8hVLKdjVSaHksdt/mmZIY0PyOOMn6DPtl2/mg3lTRZge7E0hbV5o6Iag58pM80oZTOo6wQ0KgMZjbbMdYvOiN9IoNPzZowbb7OKjNHtfzAk7MudHV33mo7+uMY9Hdry1HS1fvV4eORVfQ9CZ+osdyvOlWokPHEyHEOinlkVPaTxDdfXztpafvR08f/9FwIva5q7/zeH7fyfo3c7LuV0c4lYyBk9bf5MaNGzdu3BwLUsqYEz2GrqIbRHPduHHjxo0bNyeSU8kYmHeiB3AEuvv42ktP34+ePv6j4UTsc1d/5/H8vpP1b+Zk3S83bXDK5Ay4cePGjRs3btrmVPIMuHHjxo0bN27awG0MuHHjxo0bN52IEOKqehGjHoPbGHDjxo0bN246l6sAtzFwtJxxxhkSpVRzyr6KaopkamGqzK7IPh7b75bHIr86X6YWpsqcypwT/vt30atDuM8L97HoTi+H0yHTy9LljqIdssxadsKOxeGIuffHS2Pu/TE95t4fnfX/X3qs2zxEi+OhQoilQoj1QoiFQogwIcRFKOnhj+vbFJuFEJOFEBvr2xO/K4Qw1m/zSSFEan2742frp50thFhdv/xiIUTIsY69PXQrY6Cw8KTXdTgsWwu3cvmCy5nz4xxmfT+Ln9J+os5Rd0LG0lXHYt3BdVzy4yXM+XEOs3+YzdKspTjdXSWbcaqfF92JU/1YVNoq+WjHR1z43YXM/mE2Ny66kX2l+070sJpRf+N/C4hGCY9GA291gkHQVovjV4CLpJRDgXeBJ6SUXwLrgMvq+wtI4H1UX4P+KH2fm4UQAcD5QN/6dsf/rv+e5agmSIOB/wL/OMZxt4tuZQycyhTXFHP/svvJrswGoLKuknuX3dvtTrTOJLcql7uX3k1+dT4AZdYy/r7072SUHU/lXTdu3BwtO4p38Oy6Z7E5VUvw7UXbeXXjq1jtR+702YX8B/BoMc2jfvqx0KzFMRAF9AMW1bcjfhCIbGO9JCCtvisiqH49pwFlQC3wjhDiAqC6fn4ksFAIsRW4B+h7jONuF24Fwm5CQU0BWo2W2wffjtVhxag1sqVgCweqDpAckHyih3dcyK/Kp6i2qNk0q8NKYW0hB3MOUm4tJ9onmkTfRLSaNiX/j4qs8iz2lO5BCEFv395EeLfdX6En4nRKSmvq8Pc0nOihuOmh5FbmsqtkF3annUS/RKIt0VTXVbO7ZDe7ine1Wv6P7D8ori0mzCvsBIy2TXp1cHq7kFLuFkIMAWagnuJ/A7ZLKUcf5fbsQogRwGTgIuA2YBLK2/C8lPI7IcTpwCPHMu724jYGugm+Rl/OijuLlze8jKwPn02NnkqgKfAEj+z44WvyxawzU2OvcU2bETuDD1M/ZEnWEgC0QsvLk15mbPhYcqpyQEKYVxg6zdH96e4q3sWNi250GSGhnqHMnTKXeN/4Y9+hbsBXG7K558stpP3fjPa0UnbjphnpZenc9uttZFQo75zFYOGVSa+wKX8TL2x4gdsH395qnT4BffA2eHf1UA9HJio00Nb0o6ZJi+OPhBClwC1AkBBitJRypRBCD/SWUm4HKlCtiQF2ATFCiAQp5V7gL8DS+g6HHlLKBUKIP4H99cv7AAfq3195LGPuCO4wQTegqKaItLI03t36rssQAFiUsYgya9kJHNnxpZd3Lx4e/TBaoZ76BYIx4WNchgCAQzp4fNXj/Jz+M+d9ex7n/O8cXt7wsiu00FG+2/ddM2/EwaqD/Jr567HtSDdib34lAKXVJybXxE3PZmXOSpchAFBuK+e/O//Lz2k/A7C3dC/jI8a75nvpvbhn2D14Gby6fKyH4X4aXe4NVNdPPxb6A2vqQwIPA/9CPdE/JYTYDGwCxtQv+z4wt35ZAVwNfFHv+ncCc1HGwg9CiC2oPIG76td9pH7Z9XRh10i3Z6AbsDp3NRJJRV1Fq3nFtcUnYERdgxCCaTHTSPBNILcql2BzMPvKWudIHKw6yP6y/a445Xvb3yPaEs2FvS/s0Pc5pZNthdtaTd9ZtPPodqAbcrCsVv1fXoufO1TgpoO0laO0r2wf4V7h7CjZwU9pP3F61OncMvAWoi3R9A3oS7RPWw/hJ470J2d+EnPvj6ByBHqhPAL3pz8585Nj2a6UciGwsI1Zp7Wx7FfAV00m/QoMbrFYLjCijXX/B/zv6Ed6dLiNgW7A4szFDAgaQG+/3uwuUTkmZp2ZS5Mvxaw3s+7gOhJ8E/A1+Z7YgR4H9Bo9Sf5JJPknAeDEiUA085CMChvF1oKtzdb7bt93nJ94PhrRfueWRmg4N/5ckvyT8DH6AFBYXcjw0OGdsCfdg8IqlchVVGk7wSNx0xMZEzEGq9NKmGcYEklVXRVeei9+Sf/FtczvWb+zMX8jn838rNvm29Tf+I/p5n+q4Q4TnCBKa0s5WHUQu9NOkl8Sr214jdsH306/gH4A3DX0Lr7a8xV3L72bqxdezYPLHySvKu8Ej/r409uvN89OeJYAUwAAY8LGcG2/a/Ex+KDX6F3L9Qvs1yFDoIHkgGSWH1jO3M1zmbt5LpsKNpHgm9Bp4z/RlFbXEehloKTabQy4OTJWu5XcylzKreUARFui2Vuylzc2v8HczXP5Jf0XYiwxXNT7ImJ9YgGI9YnllYmvdFtDwM3R4fYMdDF1jjpW5Kzg6bVPU1BTwAUJF3B2wtl8uftL/v7737lxwI3cPvh2Pt75MaXWUtd6Sw8s5bzC8wjx7BL9iROGQWtgWsw0BgYNpKCmgFU5q7jt19uI9onmrqF38c62d3BKJ+fGn3tU2/8181eyKrJcn/eW7mVV7ip6+/furF04oVTU2gn2NlJe684ZcHN40srSmLt5LoszFpPgm8A/R/6TtNI0thU1htLyqvNYmr2UvSV7mR49nYlREwnzCsPP5HcCR+7meOA2BrqYncU7uf23211u8I93fowQgg/O/ICdxTtxSichHiFsL9zeat3M8mNKhu1ReBm8eHjFw/yZ8ycAu0t288L6F3hl0itEeUcRZYnq8DYdTgdrD65tNX1D/gau6HvFMY+5O1BRW0dMgAflNfYTPRQ33Ziquir+b/X/sTJ3JQCpxak8vvJxBgYObLXs3tK9PDjqQSK9IwnyCOrqobrpItzGQBdTYi3h+gHXo9PocDgdCAR6jR6ndDKx10SyKrLIqczhzqF38t7295ol9JwsT69Hwmq3srN4J/0C+xHvG89Xe75iTPgY4n3jqbHXdKiMqdZeS1pZGpV1lUR7RzM9ejob8zc2W+b0yNM7eQ9OHFU2B/6eBircngE3hyG3KtdlCAAk+CYwPWY6QeYgvtr7VbNlp0VPY3BIY+5b03Oql3cvl7cyuyKbnMocfE2+xFpi0Wv1uOk5uI2Bo6S6rhqbw9ahpL7Vuat5bcNrxPvFMyxkGG9tfculOFhYU8jEXhP5++9/p6KuAp3QcU3/azBqjOwu3c0N/W+gf2D/47Q33YdKWyUfpn7IG5vfQCLxN/nzxNgnmJ86n0UZiwAYFDSIf4/7N9GWaEprSzFqjThxYnfY8TH5uLZVVlvG29veZv72+UgkwR7BvHD6C8yIncGCtAUIBBckXsDo8KPSDOl21Dmc2B1OLGY9FbVuz4CbQ2PWmvHUexJgCmB69HSERjB381ymx0zn7Liz+THtR/WAEjWRGbEzXOuVWkt5e8vbzE+dD0CIRwivTHqFGnsNt/92O+W2crRCyx1D7mB20mw89C2FAN10V467MSCE8AXeRsk2SuAaKeXKw67UjXFKJ+sPrue1za+RV5XH7KTZzIidccRYfn51PpsLNpPon0i5rZwlWUtchgCAXdp56M+HXOWFdmln3pZ5vD3tbYI9gonyjjpqoZ2exM7inby++XXXZ7vTzraibc2e5jcVbOLntJ/xMnjxUepHhHqGclbcWXy791vOijuLKdFTCDAHsL14O+9vf9+1Xn51Pi9teInnJjzHtf2uRQhBL+9eGHXGrtzF40a11YHZoMWs15JXUXuih+OmGxPhHcHT459mUeYiLEYLz61/DoAFaQvoG9CX2wffzojQEST4JjS7oacWpboMAYEgrzqP59c9T7BnMOU2lYTokA6eW/8cg4MHMzC4ddjBTSNCiMeAP6SUizu43unA3VLKszprLF1xd3kJ+FlKeZEQwkBrzegexY6iHdyw6AbsUj15Pb/+eWrttdw08CaX4tuBigNsL9pOha2C3n69SQ5IJq8qj7UH17IqdxUXJV7USujGz+RHXnXraoHKukpG+ow8/jvWTThYdbDZ50ivSHYWt9YB2Fe2j6EhQzkz9kwMWgP7y/YzPnI8/179b4QQzE6a3WpboPIDqu3VJ2XIpcpmx6xXxkCV1e0ZcHNoauw1fLvvWxZlLOKGATc0m7e9aDvbi7bzxVlfsK90H3tK9uBp8KRfQD9yK3O5vM/l+Bh9sDlseOg9+Gn/TyQFJLX6jsyKTLcxAAh1YxBStu7AJqX8VxeNQSelPOxF4bgaA0IIH5Qgw1UAUkob0KNrnnaV7HIZAgCB5kDsTjsb8jcQ5R1FmbWMe5be4xLP0QgNz054ljpHHZsLNjOr9ywS/RIpsZY0MwhKaksI8QhpZRCEeoR2zY51MXXOOjLLM7E6rER6RWIxWgAI82yub55VkcXYiLEsP7DcNc1D58HI0JE8uvJRVyJmnE+c66L2YeqHnBF7RqttAQwLGYav0fc47dWJpdpmx6TXYtRrqbI6TvRw3HQj7E47meWZ1DpqifCMoLC20BV2M2pbe8bCPMOosldxzcJrcEonFoOFy/tczuDgwcxPnU9aWRqgvAP3j7y/zZ4FXvoTqEr4iM+ltBAd4pGyY9IdEEI8CWRJKV+r//wIUIlSGJwNGIFvpJQPCyFiUAJFq4GhwAwhxKOo1sYSeFdK+YIQ4n3gBynll0KI4aiHZ0/AiupZUAe8Ub+eHbhLStko0arG4Y/qmBiHUlq8QUq5pX588fXTM4FLDrd/HSrUFkKMEUJcKoS4ouF1hFVigQLgvfrezG8LITxbbPMGIcQ6IcS6goKCjgznhGDWmV3vI70iuazPZXyy8xNuWnQTSzKXsCZ3DaGeoURblCqXUzp5ecPLIOFvg//GsgPL+M/q/zA4eDC9vBv7Zpi0Jv499t9YDOqmqNPoeGDkA12qmd9Vx6LMWsbbW97mou8uYs4Pc7hp8U3sL1Wy3En+Sdw+6HaXhoBBa2B0+GhGhjZ6R65IuYK3t77dTJjIS++Fh84Di8GCn8kPndCREpCiwgEoj02IRwh3Db2rR8Qxj+ZYVNscGHUaTHoN1W5joNPoadeollTYKpi/fT4Xfn8hc36Yww2LbqDGXoNJawJgccZiru13LTqhng0tBguPj32c59Y+h1FrZGqvqdw7/F4+3fkpu0t2uwwBAInko9SPGBcxjmjvaMaGjyXRJ5ErUq7AU+/Z5niOO8oQaNXCuH76sfAZ6qbfwGzU/S0RpSQ4CBgqhGhQJEwEXpdS9gUCgQgpZb/6NsbvNd1wvdf8M+BvUsqBwBSgBrgVkPXrXALMF0KYWozrUWBjfRvk+4EPmsxLAaZIKQ9rCEAHPANCiA9RVsYmoOFKI1t8cVvbHwLcLqVcLYR4CbgXeKhhASnlPGAewLBhw2SbW+lGpASkEOoRysHqg5wdfzavbXwNu7RzQeIFeBm8SC9Pp8RaQv/A/pyXcB6vb3qdgpoCQjxDeGXzKy7X9YsbXuS8+PP454h/EmQOItoSjYfeg8/P+pzcqlx8jb5E+0Q3E9o53nTVsdhWuK1ZXsDWwq28u+1dHhn9CF4GL67qdxWnR51OeV05kV6RLM5cjKfBk5sH3qwSAc3BFNYoye5wz3AuT7mclTkreW/be/x18F9J8kty3fBvGngTZ8aeSWVdJZHekYR49AydhqM5FjU2h/IM6LRU17nDBJ1FT7tGtSS1KJUXN7zY+Lk4lXe2vsPfhvyNp9Y+xY7iHdQ6arlz6J30svQi0TcRg9ZAtE80YyPGsiJnBYsyF3H9gOupqqtqtf38mny89F7MiJvBH9l/0DeoL0l+ScRYYrpuJ5tzuBbGR+0dkFJuFEIE1zcsCgJKUP0KpgENSU1eKCMgE8iQUq6qn74fiBNCvAL8CPxCc5KAXCnl2vrvKgcQQoxDdTFESrlTCJEBtIxxjgMurF/mNyFEgBDCUj/vOyllDe2gI2GCYUCKlLIjJ0M2kC2lXF3/+UuUMdDjyKrIUlLBEp4//XmyKrKoddTy3OnPUV1XTZR3FPO3z2dRpnK9pRalEuIRwiXJl1BmLSO1KJXsiiYJg047X+75kvGR4+kT0Mc1PcI74qRX9mpaLtk/sD+nR51OnbOOtXlr6RvYl9LaUtLK0iizlSGlZHfRbn7L/I3fMn/jtMjTCIoMYmbcTL7a8xVzkubw/LrnXaGbjQUbeWDkAyBUuZSXwcsldXyyU13X6Bmosbk9A6ciUkr2lOxhf9l+PPWeJPsnt6lP8mfOnwwJGcJfB/+V3KpcAkwB5FTlEOYZRoR3BE7ppI9fH6od1YwJH0OwRzA6jY5wz3A0QoOzSfj7+v7XsyB9AV/v+RpQOQd/ZP/BR2d+1GX73YLj0sK4ni9QzYlCUU/y0cD/SSnfbLpQfZjAZTlJKUuEEAOB6cBNKK/CNZ0wniPR2no7BB0xBrahfoDc9q4gpTwohMgSQiRJKXehYiCpHfjObsGekj3cuOhGCmoKmNV7FqlFqWwv2s7w0OEMDR5KH/8+7C/bz+LM5gmhedV5RHlH4ZROsiuzCTAFNOuYByrn4FQj0jsSgBhLDMNChvHKxldc867qexVWu5VPd30KqJyLR0c/yi+ZvzAgaAAeOg8eX/U4V6RcwZykOeRX5zfL4QD4bNdnxPrE0se/D1f2vRKD9tRo2FNrc2DQaZRnwG0MnJKsPbiWmxbfRJ1T6UwMCR7CrYNubbVcin8Ku4p2Ee0TzaiwUVTXVTN3y1ym9JoCQHFNMQadgRc3vkisTyzDQ4fz8Y6PibYoJdCvdn9Ffk0+U3pNIdwrnNc2vdZs+8W1xewt3XtU4mCdwHFpYVzPZ6gQRCAwAeUZeFwI8bGUslIIEYGK8zdDCBEI2KSUXwkhdgEtLaVdQJgQYriUcq0QwhsVJlgGXAb8JoTojTJodgFN66Eblnm8vsqgUEpZ3tEW5kc0BoQQ36PCAd5AqhBiDSq5AQAp5TlH2MTtwMf1MZH9qFaOPYpf0n+hoEbFCiO8IthWuI2/Dv4r6/PWM3fLXDz1njw34Tk0QoNDNr8I+xp9eTfzXart1dw7/F7+teJfrhP11kG3nlS6+O2lX2A/pkZPJdYnlve2NQud8f7295tdvJzSyaubXuWyPpfhpffi+fXPA/BB6gf0C+jHuIhxzdb31nvT20950V7d9CoToia4Pp/s1NQ5MGg1GHUaau1uY+BUo9xazrNrn3VdX0BVz5RYS5gZO5Mf034EVE7A7UNup85RxwvrXyC1OJVAc6DKp9Ep77rdaefrPV/jlE5mJ812nacZ5Rm8uvFV7hp6F4HmQF7f/DqB5kC0Qtvq2qfVaLtoz1txP+qG3TRU0BktjJFSbq+/UR+QUuYCuUKIPsDK+ptvJXA5jaH0BiJQuXMNeXr3tdiuTQgxB3hFCGFGGQJTgNeBN+pbH9uBq6SU1hY3+keAd+tbIVcDVx7NvrXHM/Ds0Wy4ASnlJlSIoceyo3gHAOfGn4tO6Ag0B1JZV0nfwL6szF1JVV0VGWUZzIybyXf7vnOtF+cTh06jY2TYSBJ8EugX2I8vz/6S7Mps/E3+xPvGN0tIPFUI9gjm4dEPs6VgS7MLVwN2Z/Mn/bzqPIYED3HlCTSwrWgbk3pNwqwzU2Ov4bI+l6HX6EkrSyPOJ45Lki9xNWA5Faipq/cM6DXU2lpVMbk5yamx15Bekd5qerm1nPtH3s+c5DlU1VURbYmmzl7HP5f/01W2W1hTyEN/PsQbU94AQApJiEcIU6KnsCpnFSNCRxBtieaTHZ9wSZ9LyKzIZGXuSs6IOYMDlQc4N+Fcvtj9hes7o72jSfRN7JL9bsUjZZ/wiA90cjVBA/XJfE0/v4SqAmhJvybLbEblz7Xc1lVN3q8FRrWxnVYP0FLK34Hf698XA+e1scwjbY3/UBzRGJBSLgUQQjwlpfxn03lCiKeApR35wp7IGbFnUFxbjN1p59n1yjZadmAZMZYYZvWexWe7PmNz4WaCzEHcNOAmdpfuJtIrEpPWxK6SXXy791s8dB6MDh9Nb//exPnGneA9OvH4GH1I9k8mzDOM3KrGyJOX3quVMTAmfAzLs5czNmIsRq0Rq8PlmGJZ9jJemfQKmeWZ/Jz+M2sOrgFgafZShoUM47Lky7pmh7oBtXVODDoNOo0GBNjs6rObUwN/sz8zYmfw1Z7mcsLxvvFYjBYGBzdKCq8/uL6VfodDOsiuyGZPyR4segtDQoa4PHGgvG6PjHmER1c+6hIYWpK1hOv7X09hTSE3D7yZjPIM+gX2Y3zEeMK8Wpf2dhnqxu9uYdwBOnKlmNrGtDM7ayDdmdFho7ky5Up+Sv+p2fT08nQCTAFMjZ5KhGcEEyInUFpbqpp9SHVyNiTVVNur2VK45UQMv9sS5BHEC6e/QEpACqByCP4x/B8k+CYQZFYNUUaFjeLqvlezLm8dnnpP3pj8BlOiVFyzX2A/pkRPoaS2BLPO7DIEGliXt44Sa0nX7tQJpLbOgV6rTml3EuGph16j55p+1zA9ZjoCgZ/Rj/8b93+u86spPiYf/IytOw+adWZuXHwjhbWFVNdVE+kV6ZpXUVdBYU2hyxBo4Os9X2PWmZFS8vehf+cvKX8hxiem0/fPzfGlPTkDNwO3oMoimt7NvIE/j9fAuhMB5gDifeNpq5AizCuM37JUpntqcSozYmfw2MrHiPaJ5kLvC7EYLFTXVVNZV9ksA9eNom9gX+ZNncfSrKWsPria1ze9jo/Bh7uG3kWEVwSR3pHkVuYyNWYqD/z5AEiYETeD8xPP5+s9X/PC+he4ddCt2Jxta1l1rPilZ1NjUzkDACadllq7Ax/czWJOJXpZevHE2Cf46+C/YtQaDymTnuCbwF3D7uLRlY8SYAqgzFrG1Oip2J12zk84n1t+vYVaey3nxJ9DZV0lP+z/AcCl/9EUh3RwXvx5JPknodG4PVE9lfbkDHwC/AT8H83LAivqYxUnHRvyNrC3ZC9GnZG9pXsJMfozwiOCC+PP4ct9/3MtF+EZwf7S/S532585f5JXncfUmKn8sP8HXtzwIvePuJ89pXuwGCxttgd1o55ofk7/mWEhwzDpTASaAzHqjCzJWkK/gH7YnXa0Qsv4iPE4pRMPnQeFNYUMDBpIckAyf+b8SbBHMAODBrK5YLNru0OCh7jEn04FGnIGAIw6jbui4BTFqDPSy9KL/Op8fk77mW2F2+jnm8gQqSM4az3EToDIYYwNG8sTY59gU8EmYiwxBJmDKLWWMm/LPNe2/rvrv1zT7xosBgt1zjpiLDF46Dyotle7lrk46WKqbFXsLNnZphfCTc+gPTkDZUCZEKJVfYoQQi+lPGl6pUop2Zi/kQf+fICLEi/i8dWPu+YFmAJ4afRjxFpiWZT9OykBKYwJH8MdS+5oto29pXuZ3GsyoBJ6siuz+SD1A/yMfkyNbivScvJSbi0nrSwNu7QTY4khwBzgmldYU8jekr1U2auItcRybvw5/GPZP13eE7POzI0DbmRR5iLGhI/h9U2vu57+dULHg6Me5GC1EnAqt5WzLm8d70x7hy0FW1iavZQJkROYGjO1Q10lezo19QqEAEa91h0mOIWpslXx/LrnXRUEADMjJvBQXi6ey57DOetDvq3L5uXNb7jmJ/klMbv37FbbWpGzgpsG3MSg4EHsLtnNzYNuZnvhdgpqChgeOpz0snR8jb68uOFF3j/jffoG9u2SfXTTuXREZ2ADEIVSXRKAL3BQCJEHXC+lXN/5w+taNudvZk/JHgYGDuSbvd80m1dUW8T68r0k+CYwVtbxzZ5v8DP5tapxN2gMzcIBDZ0GS6wlbCnYcsokD+ZU5vCf1f9habbKL+3t25tnJjxDnG8cGeUZvL7pdRakLQAg1DOUWwfd2ux3q7HXkFuVi0FjYHPB5mZhALu0syJnBf5Gf77c8yXXDbiOseFjqXXUct2A67im/zVtujNPdmrqHHiZ1N+bQaehxq1CeMqSVp7WzBAA+PHAUi5PuYV+exaTXb6fN/c1L3XfVbKLtmrTo7yjmJU0C5POxL7SfTy37jnifOLwN/nz/rb3CfUMZUjwEGodtSzOXOw2BnooHbliLgJmSCkDpZQBqOTBH1D5BK8fds1uTqWtku/2fscjKx6hqLYIs87cpuxmVV0VJbZSXtv0GjlVOWwu2MykXpOaLXNJn0tcDUDGRYxzlSUC1DRxrZ3srDm4xmUIAOwu3c03e79BSsnWgq0uQwBUp8KF6QsZGjK02TZq7DUU1xZTXdf6d6usq2RnyU6X12FR5iJKraVA23HNU4HausacAaNOQ22dO0flVOVQJbXl3iqHwC4lNkfrPBundLqSd0F56GbGzqShDUiAKYDefr3ZX7afdXnrsDltzO49m5yqHED1HTmVEEKECyG+PIr1FgghfI+wzGNCiClHPbgO0hHPwCgp5fUNH6SUvwghnpVS3iiE6NEN4dceXKuS01Dx6z9z/mRG7Aw+SG1su6ATOpJ8k1x/9ADLDyxnavRUV7KOTqPDz+BLnM4HrdmPP/PW8FPaT671+3ufOvHrzfmbW01bkbOCmwfeTFZFayGwrYVbmR4znfV5jQ6mBN8EXtj3Ao+PfbxVJcfosNG8tOElRoaO5If9P5BflX/KiAsdilp3zoAbuw1yN+OJgzifOPaX7XfNivOJw7O+aZdPyACm2Ka4HlxAiREV1BRwTvw5GLQGfI2+VNgqyK7MxokyLIM8ghgeMtyVbGjUGim3lbM4Q6mvTo+Z3oU7e+KRUuag5ImbcaSWwVLKGe3Ydpe0N26gI8ZArhDin8B/6z/PAfKEEFqgRz+C/K9JUuBnuz7jyr5XUm4t54YBN7A4YzEB5gAmRk3k2fXPcuOAG5utuyhjEdV11UzpNYWsgs2M9R9EdPFBCv3NWM3h7PKJI8QcyPUh4+iTtwcix3b17p0QhoQM4cs9zQ3m0yJPw6wzE+PVWqJ0aPBQkv2S6e3XG7POzDnx57A0ayl/SfmLSx3t5/SfkVJyTvw5bCnYwqV9LsHfFMCGA3/y9uC/k1RTCzWlYPbtmp3sZtTWOV2eAYNOQ22d2xg4qXE6oWS/+pv3iQDvMNj3K/z3Enwm3svZ8Wezt2QvWwu30i+wH4l+ifiU5cKYv5Jn9GRy1GR8jb6sObiGGEsMI0JH8Nqm16i2VxPuGc7o8NF8tecrUgJSXLkESf5JTIuZxttb3ya7Ipuz48+mzllHgCmAe0fcy8Cg7pEk3X9+/1YtjLdeufV4tTC+SkrZTwhxFXABqlmRVghxJvA+SnxoFxAO3CqlXCeESEeJ8XmhEvSXA2OAA8C5UsqadrQ3DgA+rJ8GcJuUcsXR7l9H/KmXApHAt/WvXvXTtDRv69jjCDYHu97nVefxzNpn0Gl0jA8bS//A/mjQ8Oy6Z8mtymVp1lIuTr7Y1RY30iuSiVETGRsyjL9FTiP6p/vhj2cJXPAPZuPDRwHjeblKy/D9K9GW50D6UR+rHsXw0OHMiG00fgcEDuDc+HMRQtDf4MdFCRe6fsNoS7QyttY9y5UpV+Jt8OantJ8w68xklmeyNHspr258lUBzIKGeoSxMX0jfgL7cXeFglj6Y5wtLGfzVLYg9P8Mnc6Bwz4na7RNKU8+AQefWGTipqbPC5k9g7jh4ezK8NRlyNsEvD4LBm+iyfOKMgaSVpdEnoA8ZZRnEGfyJDhsOB7dSp9Xy2KrHWH5gORMiJzAmfAwvbniRans1fkY//pLyF1coL8gc5OqeKoRgSMgQXpj4Ah/N+IjrB1zP9QOuZ960eUyImoBJ17K7btdTbwi0amFcP/1YaKuF8eoWywwBLpJSTkCF0EuklCmoTr1DaZtE4LX6Vsel1HcgbOAw7Y3zgalSyiGoh/OXj37XOuAZkFIWovoMtMXeYxnEiebs+LP5Zu831DpqATBoDQz26kVF9spmXgOAJdlLeGDEAzw34Tl2l+wm1hKLyWEjrHA//PA3qFAZ7tiqYMFdeJ9+L+xaACNuBKcDvrwKbvgdLOFdu5NdTKhnKP8a9S+u7HsldqedXt69XJn9kR6h3FlpZ+bY/1DtqCO2poJ8gy82h42siiyWH1ju2k6SXxI6jQ6b08Yf2X+4phs0ev6Sm4tl80eQOA2kBOmArFWw9h2Y/gScOG30E4LV7nSJDhm07v4EJzUFqfC/JgVe5Qdg988QOQIs4Yh9vzKxrpL4AVdTVLiTAG8fei38N2LEDeDhj5CSans11fZqPkj9gGhLNNf0u4Z433h2Fe/i9U2vU2OvQSd0XN3vanTa5rcKo9aIUauiw556T7oZXdnCOKvFYoualNyPo16mWEq5rYVOT1PS6mX7AdYDMS3mH6q9sSfwqhBiEKoXwjHFSdttDNR3TLq7fqCu9aSUkw61zolESsnukt3sL9uPh96DZL/k5gIcUkLedijYSX+zHx9OfoONBVsAJ4P8+9Jn33KyZC0Wg6WZ4lZKQAoWo4UnVj+B1WHltdOeoW9ZIeT9CaUtYuHSCQ4bDLgYPALg10fAboWK3JPeGADwrCokJX8/OOoAIzSU+QUmUDP0MqqKd6oLUlBvBgUN4sMzPySvMge9Ru/qWbC3dC+zes9q1RHyol5T0ay/C+qqweyvbvw+UTDpIXDaIX05hA8Gk4VTBau9iWdA684ZOKlpea0B2P87xIyFP5RkusjZSHTq/4g+7R74vb7FjMmXtIEX4XBYGRU6ilUHVwGqAdEHqR/wzuS5JMQlEO4VjtVhZXDQYFICe5x2QFe2MG5Ju1sGN8Ha5L0DaG/DmjuBPGAgystfexTf7aIjOQNfAHOBt2ndkanb0bKVZ//A/jw34blGveyMP+HD80Fvhv6zSPaNJtnsC788pG4sM54lCsHcqGG8tPu/bC/Zzfjw0UyIPJ0nVj9Bb59Y7ux/IwMO7ID0pRA6EDwDoap5Mx384sC6FRbWN6ky+SrD4GSnYDd8fBGUZoDRGwZeCinnQlASB6SVu1Y+Qmqx6mZt0Bh4c+JLDPNJpG9lOXPHPMHzqe+SVZnDWeGnMULjxWMp1/F6xg84pZMbes1g9ME9ENoPivaBzgRnvwK5m2BNo2AKp98PY/8G+hPvuuwKauuc6LUq9KJ35wyc3HiHgYc/9J+lzi9rFfQaBT/e2Xw5e616CQFaA9u9fLh+zeNU1FXw/OnP42/yZ9mBZcT6xHLLwJvoH6p66ST6naAmQ51DV7YwPlzy/J+oUMISIUQKqt3x0XCo9sY+QLaU0imEuBIVsj9qOmIM2KWUbxx5sRNPha2C59Y916wj3tbCrWwt3KqMgdoK+OVf6uY96hZY+apy7wclw+n/hMWPqkSc+En0X/goL/W7kEptGH571yLSdjIsaRZeaPH88X51Yx99m3LRjb0Tfnu88eSbcC+YA2DNXDUIrQHOfRX8Yk7MD9OV7PtVGQKeQTDuTlj5Gqx5EwKT2Dr9fpchAGBz2nhx46u86fTHs/eZjEj7k7cSr6Q6Zx0BpQXo5HZit3/FhF4jkBotAX+8DpHDlXdnxjNQmgXl2bD27eZjWPokJM9URsMpgNXepLRQq6Ha6jYGTlqCkmDyo/Dbo+oBJHQA9J4OYYNh/2/q3GhAowWhwTH9P/w38xcq6ioAuOv3uzgz5kxeOP15EiwxBHiFnqCd6XS6rIWxECLmMIu/DswXQqQCO4HtQIdrL4/Q3vgrIcQVwM8cnVfCRUeMge+FELcA39DErdEdJYmr66rJrsxuNb24tn6otkqVhTvyxvqbd/3uFOyENW8pa7s0E6yVkL8DT70Zz9/fgoGXgGcQIfuWQtggSDoDlj0HHoFQvA92/QhjbgckBCbD+vnqZnTDH1CZp9zYgT3a4m4/RfVpJEOuhCVPqBwKgLJMCotbJ/hlVuVQ7ROB5zc3wuSH8XY68S49AAGJkL0Wosfi7x0Cfzyjci/6z1HhgDVvQfJZyuho2ftBOqG29PjuZzfC2qRLoUGnpdrmFh06aSnNhB/vUOeCTxQMmA0bPwajBaY8Blu/gINb1MNK2CA47R7qcjezW+Q328xP6T/hoffgkTGPnIi9OC5svXLrJ/3n94dOriZooGkLYyllOvWtiqWU76OqBxqoBS6XUtYKIeKBxUBG/bIx9csU0rzV8bNN3l/V5H1b7Y33AAOafP4nx0BHjIEr6/+/p8k0CXQ7Sb1AcyBnxZ3FJzubH3uX68srGAZcom4mditodDD8WgjuC96hykvQ+wzwCgG9B6x6Q7mh05bC8vqWnvt+A/84GHYNbPoUhvxFLffHM2p+9FhAqpM2eWbX7Xx3IX6KelIXNBoCAHU1JGta5vbAOeHjCBDBysuiN6kbud5LGRIN+ETBrA/UBbA8F9J+h8yVKk5afkAZBFUFjct7BYPvqaPt0NQYMOo0FFSeNErhblpSvB9ixqvrTEA8fH2Dup4BpH6jrlclaWDwUNU1vz+JyezHeRNubuaVA1Xye7JRf+M/0S2MPVAhAj3qSniLlLLtjmrdgHaXFkopY9t4dTtDAECr0XJ5yuWcG38uGqHB3+TPk+OfpG9AvUymRgsjbwDfGPX5tLtV7BkJq99UN/DtX0Pqt3DpF6AzQmkabP28+RcV71eWd2CiclU3xT8OyrKV1+BUJHo0nPGkMqZaSJz22/YDT4x+FD+jH1qh5fyoyVwcOg7N6tdh6VPw833KQPP0h6iRjSuWZakL3ML7wC8K8rerJyGhhegxcPGnEFZf5xw2SH32ba1pcDIipaSuaTWBO2fg5MLprA+H1YueeYUrY3fDfMhYqUKeI2+CMX9V155tX0Kfc2D7t+r91MdAo2NyYQ7XJM7GoDHgqffkH8P+0Ur5003nIKWskFIOk1IOlFIOkFL+dOS1ThwdqSbwAO4CekkpbxBCJAJJUsofjtvojoEo7yj+Nfpf3DjwRoxaI8Eewc0X8I9VyXwVucqF338OVBeq7P/lz0FgkopL26pg+pP1ZWpttMMVGpWktvjRxmneYSovwOgNwX2O4152Y8y+MOpmlTdgt6n4fT2mAXM4J+E8RkaMxZafSuiGj9AXf62Og1+s8tL88oAyyhKnwen3we//p1Y2eCqjIG2Zioue+TQkTFYXRoArvoPqYpVcdQqJD1ntTnRagabe8DLq3dUEJw3lubDuXVj5ikqWnfQv5Q3Y8jkE9VHer4GXwvr3lKezPpzJjh9g1K0QEAf2Ohh4GcH9zuOvgcnM6n81Wo2WMM+wE713broJHQkTvIeqgRxT//kAqsLgiMZAvUrhOlTSxVkdHeTRYtAaiPI+zJOhhx/0uwAW/AN6nwkrXoK8bWpewU749TGY9b7KjA9KhH4XqVhcA74x4J8AP92rYnYRQ9QNyuyvvAkhfU+p0rZW5GxS4RQPf7j4E+Up8e0FoYOgcBch+34Fr1DoN0vlbgAMvgwWP6xCAaASM5129eRfnKYMLICc9XD512p7OkPjd5p9TykjoAFrnROjrjGZ2KTTUuP2DJwc7Pwe/nhava+rURUD575aP9MJngGw6MHG5de9A1MehcTTIGslvNtEIrj3FLQGM5GGyC4bvpueQUeMgXgp5RwhxCUAUspq0VaLq7b5G7ADOLF3RilVYltZtrKc9R7ww50QMUzVqzcYAg3UVUN1ERg94Nubldtt7N8ge51yX8eOh5VvQPIMSDkPIt3uNhc5m+C9M9VvCMqDMu1x+O+lcO7r6sk/MBG8wyFzhfpt87apJxtni5vYvl/VOuU5UJKhpsWMV+5QzanZlKgltU00BsCtQHjSYKuCDY09UvAMgiFXgFeYOqeEBtL+aL1e6rdgsMDCJgn0Jl+Vd+PGTRt0xBiw1Zc1SID67Ejr4VcBIUQkMBN4AhVmOHHsXQyfX6FuUEKjbjBpf6iny9pyZRy07JBXW6YSA3tPV244kw8Ep4DDCvGT1MtNa3YvbP5bSqealnKe0lKvLoLY69QTj5QqDNP7DNDqW2/LKwRK0pW3ZfOnSq/AJ8ptCDShqRQxqARCt2fgZECjrk8Ht6r8mHF3wtYvIbgfTHoQNn6kkp5b4h8PEYOVuFl5jvr//DdPjbJmN0dFR66mD6NqGaOEEB8DvwL/aMd6L9Yv12YzIyHEDUKIdUKIdQUFBW0t0jmUZsG3NzXeoKQTyuo1KDZ+BLt+gAktKjOSz4L9S1QpYlB97L+2TGWwF+3nZKNTj0VbLVSjRoJftMrNaKChHnrVG+p39u2lnvobBwWTHwaPYPU+5VylExE+6NjG183p6LGorXNi1DY1BrRuz0An0WXXqDa/HEicrnJlBl6s1DwD4sBWrs6ZpDOVxoB3k9i/wUt5LXf+qMKfl30F1y+B2JOvasBN59GR3gSLhBAbULWOAtU0ofBw6wghzgLypZTrhRCnH2K784B5AMOGDWsjQ6+TqC5qrQ7oH69Ooopc2PWTKkubNV+5q3VGOLgNdv6ghIlslc3XHX7tcRvqiaJTj0XSmbDqtcabvU+kcltSXxet1UPBLogcpsIuoCSEy7Jh7B3qwqU1qOUqClQoJn+bMuL6z4agk7tdcUePRUvPgMHtGeg0uuwa1RZ6swqlnfuaSiSsq4aABOWhdNqVmJfOqBIFPQNVArTdChq9EvmylkNIP0ic0qXDdtPzOKIxIIQY0mJSbv3/vYQQvaSUGw6z+ljgHCHEDMAEWIQQH0kpLz+64R4DXiGNN/4GHDaY+ICKSedtVyeNyUclvO3+RWW3j7tT1bv7Rter3gHj7oBeYw71TW4AIoarJ5LlL4C1QnldSjNg4b/AEqFKnXYthJTzIXyo8sCE9lfH4Kd71BPQpAdVOCdmNCBVXsEp1nyovbQ0Bkx6d2nhSYPBDJ9dqq5XAGY/OPslGP93yFwFRXsAqc6Z7d8qoa+8bY3eOf/YEzVyNz2I9ngGnjvMPAkcMmgupbwPuA+g3jNw9wkxBJwOMHrB7A/g0zmq9ExvBr0nfHGFugEFJavM98p8VSFgt6oEtf1/wDmvqHr1PjOVT0TfWjTHDeCwg71GZfzrjarkL3qMepqvKVcVA1Kqp/+F9yuDyicKitLggrdhxcvwa32JZt8LIG4ifP83WPSQevo54ynlKtW3t4/HqUNtE40BUGGC2joHUkran+frptthr4OVrzcaAgA1JUpIyF6rBNICk1STogn3KoGznQsgrJ8ypMfdpTQ33Lg5Akc0BqSUE9uzISHEVCnlomMfUieTt11J1matUk+hV3yvXP4eAY197/O2NVYSlGbChW9DcG/1BBvSH7yCTtz4ewq5m2HVXNUsqP9s6Hch+PVqvHFbK1WopgEpVbOopDOVuNC3N8Kw62DQpSo84BcDH10AhbvV8nYr/HCHEhWKaOmsctPSM6DVCLQagdXuxKR3e1N6LLJOqWu2pHCXChsMmKOSA9P/gK+vU144gIiBcMsqVf6sP1wvHTduFB2pJjgSTwGHNAaklL8Dv3fi9x2Z0iz46CKoqFftyt+hbloXzFOeAo1OudxqShrXiRgGO76H1P+psMK1v3TpkHskxfvhg3Mbf8dfH1ElnGc9r57oAbxDYPh1zeWFhVCJUb/U10gv+Lvyxly1QPUUaDAEmlKa6TYG2qC2rrFJUQNmvUoidBsDPRi9h/KG5W5qPj1+shIgihwOS/4Dmz5qPn/3LzD+brch4KbddKYx0P18kfmpym098kblui47oGSGS9JVJzv/WCVc8+ujykiIGQ9hAxpvWBW5SnDItzPaYJ/E5O9oblABbP5EaTI0JPqV56oyTP849XsbLcqt+dtjzdcr3q88NkFJKoRQ1kLmuWnWtBsXLT0DAEa9liqbHT9PwyHWctPtqS1XYbUzn1aaA0KA0Qcq8mHcX9UyUcNbGwNxpytjwY2bdtKZxkDXZtkeicI9UJGnkmz+fElVCgQkqMQ1bZOLY8QQmPm8uqEtew52fNd8O9rO/IlOUrRt3Gy0BvXbSQn7lsD/blbxTb9YmPwvFeMUou2nf61eeRLOfR3+e4kK6wgBE+6DkJTjvjs9kdo6J3ptc3vcpHcLD/V4NFqV/Lx/KexeoLyZgy6D0IGNy8RNhJjTVKgAIKA3DL68VU8QN24Ox8l5p7PXqSz26LHqKdRR372taK+Ka/c5t/nyAfFg8IZeIyGnSXFExNBGfQE3hyakr/qdCnY0Tht/t4pXFu2Fzy5r1HcoSVNhgT5nq9LDgZcqL0IDcROVVwAg7jS4cakKDZgDlJfBnTzYJm2FCUw6LVVuY6BnY/BUxvCuH9VnRx2sfx/OfaNxGb9omP2+8mI67fXKnm0IEblxcxg60xhI78RtHRvVBSrubwlrNAQaKElTJYOWFu5m72CVeRs1WlnYoQMgdoJ6QnVzeCzhcPHH6uklP1W5KHuNVgqBpZmtVR0rDqoSzowVMPx6VXWQuRLChyixFA//xmUDEtTLzWGpqXOgbxEmMOk1VNvsJ2hEbjqFqiKlddKSjGUw+NLGzx4BqlOoGzdHSXt0Bi443Hwp5df1/x92uS7FaIHgvqBtkjzjF6Pq1LV6deK0hVcw9D1Xvdx0jIB49WqJZ6ByV8omUSS9WcUzR1yv5keNhv4Xdd1YT0KqbY5mCoTgViE8KWi4luVtV5+FgLhJkHSWqs451LXMjZsO0h7PwNmHmSeBrztpLJ2H0QumPQpLnlRNPWrL1RP+unfUfIOnmu7pLhk87gQmJ15n6QAAeD5JREFUwaSHVAdIUBezqY+CtUoJpKx7DwZdDuPvUu5ON0dFja11AqFJ7w4T9Hh0ehh2tRLlqilW0tx7FikRooAEle8Ue5o7P8DNMdMenYGru2IgnU7USJjxtKrRLU5TNeoN/PqYylQfMPuEDe+UQW+CETepBKeKXNVutaZMJQY2sOF99QQ09VF386GjpNpmJ8CzeRmZQaeh2uoOE/R4osfA5d+o8NrSJ+FAvXx30V74ZBbcsBSC3blNbo6NDuUMCCFmAn1R0sIASCkfO/QaJ5jARPVa8UrreVs+cxsDXYXRU5U/NbDqjdbLbP4YxtzqTnw6SmraKi3Uaah0GwMnB2H9laHcYAg0YLdC0T63MeDmmGn3Y5gQYi4wB7gdpSkwC+gZft3gvq2nhfTr+nG4UXgFt57mH6fCN26OimqbA2MbYQJ3zsBJhMFLedBaYvbr+rG4OenoiE92jJTyCqBESvkoMBroGa3jBsxunmjjEQD9Z5248ZzqRAxT1RoNaA0w5RElDOXmqKixOTC2UBo06jRUuD0DJw9+0XDmU82n9Z/l9gq46RQ6Eiaoqf+/WggRDhQBPUMOLrQfXLMI8raqzyH9INBdrnbC8IuGSz6Fg1tUImFwH6VV4OaoOZRnoLL2BBoDpVkgHaqSx03n0Pd8CEiE4n3Kwxbav3kprhs3R0lHjIEfhBC+wDPABlQlwdvHY1DHhcB49XLTPfCJVC83nUK1zd6mMVBSZTvEGseZFa/BH0+r9zOfc5eOdhZ6s8q/aZqD48ZNJ9ARY+BpKaUV+EoI8QMqibD2+AzLjRs3HaG6jYZEJr2GyhMhOrTxI1j5Kpz1guqit+AeSJwGpjbi3W7cuOkWdCRnYGXDGymlVUpZ1nSaGzduThw1bYQJzHotVV2dM5C1VslNT3pI6Xj4x6n8kE0fd+043Lhx0yHao0AYCkQAZiHEYBq7E1oAj+M4Njdu3LST6joHZkNLz4CWKmsXVhNUF8MXV8Co28A3qnF6whRlDIy6uevG4saNmw7RnjDBdOAqIBJ4vsn0cuD+4zAmN27cdAC7w4nd4WzVqMis13atzsBP/1SVIr1GNp8e2h+WPav6VLjbgbtx0y1pjwLhfGC+EOJCKeVXXTAmN27cdIAqqwOzXotoIUlrNnRhmODAeiWZe+7rredptBA5DPb8AsOvc02uc9ax7uA6siqyKLeVU1xTjElnYnrMdJL8k7pm3G7cuAE6ljPwpxDiHSHETwBCiBQhxLXHaVxu3LhpJ5U2Ox6G1na9Wa+lqqsSCH9/SlUMHKrFdNggZQzUk1eVx0XfXcTTa59mWfYy9pTsweqwkluVy3W/XMfXu7tfyxM3bk5mOlJN8F7964H6z7uBz4B3DrWCECIK+AAIQZUizpNSvnR0Q3Xjxk1bVNTW4WHUtppurlcgdDolGs1xbGRTnAZZq1UXykMRNhDWzAOHHadGw12/38XAoIGcFXdWK4/G2PCxPLX2KeL94hkYNPD4jduNGzcuOuIZCJRSfg44AaSUduBI2Ul24O9SyhRgFHCrECLlqEbaQXLLasgoqsJqV0N0OiXZxdVkF1fjdMojrO2mp1JabSOtsIqy6kPX1zudkqziarJLqpGy5/8tVNTa8dC3Ngb+v72zDo+rShv474xmksnE3aXu7hQKbYG2lELRRcribruw8LHI7rKwuCyyuHvRtlCgVKDumjZptHHX8Zn7/XEnaaZJ26SN9/6eZ55kzj333vfOufLe97yiUgn0GnXnhxdu/wiST5fLUh8NQxD4hkLxTpbnLMfsNDMneU4LRQAgwi+CSwdcyiNrH8Hl7n3plItrrF73HoB6q4Oc8gYq6m3dKJmCwtFpj2WgQQgRgvyGjxBiIlBzrBUkSSoCijz/1wkh0pAjE/admLjHx2J38tOeYv65NI1qs50LRsdy47QkftxbzKurMgG4+fQU/jQhgVCj/jhbU+hNbM+r4qFv97C3sJYh0Sb+df5QRsV7520vrbPy4fpc3liThUoIbp+RyiXj41pU/OtN1Foc+Opbv5SNeg11VicmH23n7FySYOdncNpfj983YghSzlrerljDnOQ5qMTR30XGRY5jZf5KlmYv5byU8zpQ4M7DYnfy455i/rlkHzUWBxeMjuWOGamY7S4e+2Ev67MqSQjx5fEFw5iSEtKqIqSg0F20xzJwD/A9kCyEWIts/r+9rSsLIRKBUcDGI9pvEEJsEUJsKSsra4c4rbMrv4a7v9hJZYMdtwRLdhWyOqOc537JwOpwY3W4ef6XDH5PP/l99TU6eiy6ksJqC9d/sIW9hbUA7C2s5foPtlBYbfHqt+pAGS//dhCb043F4eKp5QdYf7CiO0Q+Ju0Zi1qrAz9dS8sAgJ9eTa3F0RkiyuRvkR0Eg5OP3zdiCPUHllBlq2JY6LBjdhVCcF7yeby247Vutw60dSx25ddwzxc7qTI7cEvw1dZ81h4s5y9f7mR9ViUAuRVmrn1vMwdL67tKfAWFNtEeZWAf8A2wGSgB3kT2GzguQggjsBi4S5Kk2ubLJEl6Q5KksZIkjQ0LC2uHOK2zv7jO6/vgKBMrD5S26PftjgJyyutZsrOQlftLKK624HZLZJTUsWp/KXsKarA6Tq0iLx09Fl1FVYONvQU1lNd7Tw2U19s5VGlu+u52S3y19VCL9X/YWcihHjZl0J6xqGpw4HtUZUBDTWcqA3u+goQp0Ja33Igh6Aq3Myli/DGtAo0MDB6IQWPgt0O/dYCgJ05bx2JfUW2LtrJ6O3sKvdttTje782vIKqvvUeecwqlNe6YJPkDOLfBvz/fLgQ+RSxkfFSGEFlkR+FiSpE53EU4K8eX605Iw6bWU19vIKKllYKQ/aUV1zBocAcDP+0roH+HPrZ9sb3qTnDU4gisnJXDd+1uwOd0IAffNHsCiyYkYWvHUVuh6KuptuCUJX62GWpuDIIOW/GoLy3YXIYRAJWBCcggDI/0pr7MR6KvDZDhsHlepBAMjTWSVNTBzcAQuN/y8r5iEUD/+9OYG/r1gGFP79R4lqJEqsx2/o0wT+OnUVJs7SRlwu2HvN3DmI23rbgikUsCZPpFt6i+EYGbCTN7Z8w4zE2aejKRdQpCvjuRQX/55/lC0ahXLdhXio1Fh1Gta5HvIr7bw1PIDvHL5KOKCfQk3HcPfQkGhC2iPZWCoJEnXSZK00vO5HjhmqTkhT4q9DaRJkvTcsfp2BAdL61idXkaEvw/f7ihkRVopV0xKYkx8MH+aEM/KA6WsPFDKnybEMyU1hAabE3+9Bp1axc/7StidX4PN6QbkqdD//HSADMWc1+3UWux8vjmPy97cwE97ivnze5u48LX13PPFTlbsL2XpriKCfXW8ccUY1CrQqVVEBviwdHcRN320lWW7i5qsPGcPieCScXGsSCvl94wyFk1OZGJyMLVWJ3/9aheltb2v3EZ5ve2oPgGyZaCTihXlrQedsc2JhLJrssn28SOlqrDNuxgVPopScym7ynadqJRdxoAIP+48qz//XJLGPV/sRKVSMS4pmL/O9s6ZMH9ENHanm0vHx3H7p9uZ89IfvLs2m4oGxblQoftozyvvNiHEREmSNgAIISYAW46zzhTgSmC3EGKHp+1BSZKWtVvS45BWWMvfvt7F7CGRPL4sjYGR/vx19iAe/GY310xJ4uXfDgLyg8ItQXaZmQWjYogN8qXabKfK4qDO5kSnVmF3uZu22xsfDn2NjdlV3L94N7eekUpmaQOTUkPpV28n0uRDUbWF5DAjD3+/l7euHssFI2PIrbTw4ooMACob7Nzy8TY+u2ECJbU29hfXodeomDk4go835vHCrxn8Y/4Q4oIM5JSbyasy8/O+EqwOF+MSgxkWE9AUlpdZVs+WnEqqzQ5GxwcxIi4AnaZ183xXUlZnY0h0QKvLjHoNlQ2dZBnY9RkkTm1z920lWwkKScZUsI2SEW2rYqhWqZkRN4N397zL82c8f6KSdiqZZXVsz63GR6viYGk9Zw6KwGx3Eu7vw6asCmICfXj5slEU11gJ8tVisTvRazXct/iwgvPYD/swGbRcOFqp5KnQPbRHGRgDrBNC5Hm+xwMHhBC7AUmSpOFHriBJ0h8crmXQ4ZTWWsmrNONwubE5XJwxMJxIkw83T09maEwgNqebu8/qBwhum5GKzeHG30fNp5sOUVRz+CH/5AXDsDndjIgL4NNNedjNsjKgEhDm33u9zPsCpbVWvtmez1mDwhkWY6Kw2sKW3CpWpJVy0dg4hkT5c+bACOYOi6KgyoLd6WLJTu83z37hRlbuL+N/a7Ka2sYmBHHeiGi+31nI+qwKdGoV987qz5/f3Uyt1UlskIFqs4PSOiv9I/yRJLj8zQ2U1Mpvb0LAW1eN5cxBEV36e7RGaa2NKamtWwZMBi1ldZ2g0NobYO+3MO+lNq+yrXQb58ZMxrj5I3C7ZMfDNjAtdhoP/P4Ah2oPEWeKO/4KXUhZnZW3VmcRavJhSLSJfYU1DIsNxN9HQ05FA/0jjAQb9by6Kos/DpYDcN6IKDSqlkbZjzfmMm9EFDp19yuYCqce7VEGzu40KU6AtKJa/rVkH9P6hZIQ6suH6/NYlyl7hU9OCcHmlFiTXsb5o2J48dcM7C43AQYt/5g/hHqr9/zdG2uyuHx8PI8vS+Pus/rx7M/puNwSN52egsOlOPh0BA6XG7PNib+Pts0JcLbkVPLId3u5ekoC32wv5KaPtgEwJj6IFy4Zyb+WplFvdbImo5wf9xQD0D/cyDVTk3jo2z1N25k1JJI31mR6bzu3irv7hQIQ7q8nNkB++NdanSSE+LJwTCz/9UQd+Os1PL5gmNe5IEnw9PL9jE0MIsCgO6nf5mQprrUeNTQywKAlp6Kh43e66wuIGAJ+oW3qXmwuocFuJiQ4FadPAH5lB2iIaFvKEYPGwPTY6by15y0em/zYyUjd4WSW1hMb4sebv2fy7/OHEeCr48UVGUgSJIT4MiTaxM78Gny0Ki4aG8uXW/JZn1XJwjExLbYVH+SLpg2OlQoKnUGbzzxJknKP9elMIY/EYneRXlzH+KRglu8rYVd+TZMiALAuswIh4ILRMTz784Ems3+NxcG/lqaxYLT3hVhtcVBlsTMkOoBaq5N/nj+UKycl8PGGPOqsp1ZEQWdwoLiOh77ZzXmvrOXxZWlklR3fD6Oi3sZfvpJDRGstTtY3G1+jj4b31+dQXGslKdS3SREASC+tZ3d+NYMi/ZvafHXqVpU6lyRh8tEwJNrEoCgTVZ5ERQtGxfDSiowm/5E6m5O/f7eHC444byobHDic3ass2pwuqs0Ogv1aV0hCjXoKqzvYMuBywB/Pw6C2x/9vK9lGalAqKgQNYf0JOHS8GUZvzko4i19yfiG/Lr+90nYqDrfESysycDjBbHfx9bYCGgMEcivM/La/lBA/Hb+mlZIaZkSnVlFWZ2N8UggBzZxbfbQqrp6c2LmZIhUUjkGvVEPrrQ4OlNTx/K8ZqIRgR17L3EdpRbUgwZHJBsvqbF4XIcCcYVF8t6OQL7fIccFWh4vXV2dhdboI9NXibOZDoNA+imssXPv+Zj7fkk9uhZm3/8jmr1/tOmaGQICKejs55WYmJAezPss7D8CgKH+25FQR7KujoJUH3drMCm4+PYVbz0jlrrP6EWrUMTE52KtPkK+WodEmrp2aRJ3FyV++2smgKBMATrfUQnmosTgwHJHl75opiYR28zRSdnkD4SY96qM8RCJMPuSUN3RsCNvG/8kWgchj5wpozubiTfQLTAWgIaw/gbkb2rVLf50/Z8afybNbnm3Xep1NndWBzelGq5EtNEeyLa8anee8ya2QxyrAoCXQoOGrmybx3MUjeHrhcL6+ZUqLBFkKCl1Jr1QGaqwOPtssx4vnlDcwMMq/RZ+BkSYQLcOf5QdDCCPjAogJNHDNlESsThf5VXJymg1Zleg1aqb1C+X2Galc/c4mL6uDQvvIKmto+m0b2ZpbRU6F+ShryAT76YgLNlDZYGPoEc5xB4rrGRUfSJXZTnRgy5CskbGBvL46i9dWHeTFFRncv3g3l46LZ8GoGCJMes4YEM4zF43gzd+zUKkEP+0txi3BxxvzePKCYQT5atEc8XA1+WiYmhrK6PhAYoMMPHjuQC7oAc5eewpqSQzxPeryIF8tGrXouKiY4j2w5mkYd0ObVymzlFNqKSPeJEcdmEOS8anOR2OuateuZyfNZnf5bn7L6968A82JNPmgU6twOCE6sGWRptEJQWzLkY8zKcyPWYMieOai4TTYXfSL8OeC0bFcNDaOwR5FVEGhu+h1ykCtxU5JrRWjJ666osGOXqNieOzhB8aouEA0KsH3Owu59fTUphu7n07NTdNTkNwS4xODeOKCoSzbXcSXWw6bHtUqgdstYXO6+eeSNGqtTv69LK3zwrP6ODpNy1NMiNbbmxPqr+eZhSMob7ATE2hgdLO3psoGK9dNSybIT0d+lYUZA8ObliWF+jIpJYR9RbW4JZpMtnsLa6mot/HaFWN47uLhmAwawv0NvLYqk82em3VGaR35VWZyKxp49LwhaNXyeeOjVfHcxSOZkBzCh9dO4PvbpnDDaSlE9IDY8HUHy+kX3lIZbkQIwXkjornl420nb+GqyYdPLoJx10NAyznvo7GhaAMDggagFh7LikpDffhAgnLWtmv3erWea4deyyPrHuFA5YF2rdtZDI7055F5gxFCVrzOHxndtCw6wIc/TYjngw25TOsXSrCvjgCDlus/2IroPL9qBYUTotdl09mVX8u9X+7kzjP78eA3spPY66uz+NOEeK6alIDLLZEaZqS0zsbo+CDK623cPiMVh1vC7ZYoqrEyNNrE5NQwfs8oJzrA0OQhDnDVpATeXpvdlIwI5FS3FrubgKNUZ1U4Ov3CjZw5KJwVaYezQF45IYHkUL/jrjshOYT//WkMacV1xAcbOK1/KJIElfU2KuptzB0ehVGvYXxiEDMGhhHqpycmyIe8SgtCHFYETAYNk1NCmJISTEKwL0F+euKcbnLKGzDbD6e6vXx8PD/sKiK3wsz/rhjD0jumUVZnIyrAhySPvH56zVET/HQ1Dpeb3/aX8o/5x0z3wVmDIvg9o5yN2ZVMSW2bw18L6orhvTkw4FxInt7m1dxI/J6/hrMTvf2P66KGEZL+K2WD57ZLjNSgVC4deCnX/Xwd9469l3nJ81C3MSqhMyiqs/P+uhyuPy0JH62a6EAfXrx0JDanm1Cjjnqrg3tn9md/cS2+eg3Pr8hgTEIg/SOOrsApKHQHPeOu1g4ySusoqbWxJaeSFy4ZwZ7CWoJ9dUQG+FBUY2VXfjX3L97Nq38aTUyQgc05lfjo1JTUWIkN9sXqcDEwyoSPVoVeo2JItIn55mgKqiyMSwomzF/Pe+tyvPZ5ybg4wpUQwxMiwFfHP+cPZd7wSvYX1zIiNpCxiUHoW6my1xpxIX68+FsG/SP8Ka2TMxD2i/Cn1mIn3F9PToUZu9ONTqNiWHQAw2KDSAkz8s7VY1mXWYGfXkO/CH/+9vUuSmptPHjuIG44LZnIAAOvXTGGDVnl7MqvISLAh625VeRWmNGoBLFBBvpH+Pfom/b6zArCTXrC/I9toRBCMCwmgLUHy09MGWgolxWBxNNg8PntWnVP2W7UKg1Rxmiv9obwQUTuWoyutgi7Kapd2xwfOZ5wQzgf7vuQ13e+znXDruP81PPRqLr+dlZYZSa9tJ70XzJ4b9EYJiaHsiu/mmqzg4DEIEpqrLglicvGx7Mrv5pnLxrO+KQQJWRZocfR65SBGM+83NfbC/l6eyFzh0dSZ3bw/vocyupsTQ6Dy/cW8+Klo7A6XTy5LE22CFgDuHfWAAJ9Zc/rSSktb4x2p4v/XTGGfyzZR1mdjcvGx3PVJMXL92SIDjRw/qgY5IKV7ad/hD//XrafAIMWlYAqs4OXLxvFpuwK1h0sx0+v4a+zBzA4Rp4q8tVrOWNgBG4JHvpmD0XNHLteWXmQ80ZEExngQ3yIL/Eh8UxKMfPSigzWHiwnJcyPR88bwsBeMIf7454ixiYEH78jkBpuZPWJFOdqqID35kLsOBh+cbtWlZD4LvN7xkWMbWEUl9QaauLGEL7nO/In39RusRIDErl/3P2kV6XzZfqXfJL2CU9Pf5qUwJR2b+tkCDHKzpsqAbmVFsx2F7+nl2PQqqgy2xkSZeLqiXEEGvW9MtW1wqlDr1MGRsQFMHNQOL94zM4ltTaiAoSXqR9Aq5bnpMcmBPP2ovHUWR2E+OkxHKWgSyM6jZpZQyIZmxCExekmwl+PRt3rXCv6FDMHRbB4az4HSmQnuIQQXwZG+nP20Ehumm7FR6smpJVy1A02l5ciAKDx3LibExfsy+MLhnLXWf3w1WkIOkqYXk9CkiRWpJVy3+yBbeqfEmbk1VWZuN1S2xXb2kL44HyIGgEjLm+3jJuKN2F2NjAwZFCry6sSp5D4+0sUj7oMp6H1DIrHQgjBgOAB9A/qz+8Fv3P1j1fz+NTHmR7X9mmMkyU5zI+/zxnE48vSKK2z88G6HM4ZFkWgr5aV+0vZlFPJgh7gaKqgcDx6nTIQYTLw1MLhHCipx2xzkhxupKTWypJdRU1WAZWAi8cevgADDNoW4YTHI7iVh4tC95AUZuSDayeQXlKH5JkmiPI4cMQEHd2Tfki0iUBfrVehnrtm9mu1KIxOoz7mtnoaaUV1qIRoNZqiNUwGLf4+Gg6W1bdt6uPQJvjiSuh/Dgy5oG1VCZtRYi7lo30fsyD1fFRHcZZz+gZTGzOK2A1vknPGX9q1/eYIITgt9jRijDH8fe3fuXnEzVw26LIT3l570GnUXDo+njEJQdSYHXywLocvthyujPn8xSN6hXKpoNDrlAGAID89E5MPP6xjAg18dsNEvt1RiAo4b2S0ErPbx4gw+bTbez8l3Min10/kxz1FZJc3MHd4NBOTQjpJwq5lya5CxiYGIdrxkB4U5c8fGeXHVgZs9XLo4LYPYOKtED+h3bJl1mTxyo7/MjVmCtFH+AocSfmAWST+/iKhacsoH3Ruu/fVnJTAFO4ffz//3f5fdpXv4m/j/0aAvv0Wh/bio1UzLDYQgE9vnMiPu4rIr7Zw3ohoJiS1bRpHQaG76ZXKwJFo1SrGJ4Uwvo/c6BU6jkFRpqZkQn2FzLJ6PtqQy0NzBuM6MqvWMRibEMz763O4aGws/j5aOZOgrQ4sVVBxEA7+Cjs/h5hRMOc5MATJNQSOglNyYnPZsTmt1NhqOVSfz9aSLWTXZDEj/kz6B/bD7Tp2kSS3SsOhsVcRt+kd/A9tpWzgbCyBcTj1/ri17Q/fCfEJ4W/j/8bijMWcvfhs5ibPZUrMFBJMCQToAzBoDBg0nRcWNDQ6oEVeDAWF3oDo0MxkJ4kQogzorNTGoUB5J227I+gK+colSWpTjYmTGIue/jsfjx41DuA9FiqDSR13xycjASR3O/MGCBBCRf3elWUVS57N23KD34AxUWrjkd3arGB0oFNta54802OjqTiJoj3iKPIduOfATkelozHP+AmPRTN6+zl/NLr6uNo1FgodS49SBjoTIcQWSZLGdrccR6Ony9dWevtx9Hb5T4TuOOau3mdn7q+vnjN99bgUWkdxk1dQUFBQUDjFUZQBBQUFBQWFU5xTSRl4o7sFOA49Xb620tuPo7fLfyJ0xzF39T47c3999Zzpq8el0AqnjM+AgoKCgoKCQuucSpYBBQUFBQUFhVZQlAEFBQUFBYVTnB6lDJx99tkSoHw679NmlLHoGeMAylh08qddKGPRc8ZCoWPpURkIy8t7R94Oq9NKdk02ZqeZOP84wn3Dm5bl1+VT3FBMoD6QhIAEtCrvmgjHW95T6C1jAVBpqSSjKgOry0pyQDJxprimZVanlZzaHOod9cT7x3uNVW+hN43FyVJYX0hhfSEmnYlI30gKGwqptdfiklwkBSQR6RfZrfKdSmOhcGrRo5SB3kC1tZo3d7/Jh/s+REIi2i+aF2e8yMDggWwq2sTdq+6m1l6LRqXhb+P/xvkp56PXyHUUNhdt5q5Vd8nLhbx8fup8fDTty7mvcJjsmmxe2fEKy3OWA5AUkMS/pvyL4WHDqbHW8M6ed3h377tISET4RvDyjJcZdJQqegrdy47SHdzx2x1U2apQCzXXDbsOH7UPL+94GbfkJtQQysszXmZo6NDuFlVBoc/Ro6YJegN7K/bywb4PkDxWrcKGQl7a9hKF9YU88McD1NprAXC6nTy+4XEyqzMBKDOXeS+XnPxr47+aliucGNtLtzcpAiArB5/t/wyb08a+yn28s/edprEqMZfw7JZnMTvM3SWuwlGotFby8LqHqbJVAeCSXPxv1/9w48YtyWmXyy3l/GvDv6i11XanqAoKfRJFGTgKZoeZamt1i/aC+gLUQs3UmKnMTZ5LmCGMLSVbKDWXUmou9eorIVFsLgagwlJBibmkxfZaa1MAi9NClbWK1kJfbU4blZZKXG4XByoPtFi+rXQbFZYKihqKWizbWrqValt1Z4is0A6qrFVeSlm1tZrsmuwW/Wwum9f3vRV7qbHXdLp8J0tOeUO7ikgpKHQ3yjTBEbjcLraWbOWVHa9QYi7hkgGXcG7SuUT4RQCQaErk3rH3sjxnOdk12cxKnEWYIYxw33DCfcNbKASRvvIcZ7BPMBG+ES0e/hG+EV1zYL2IHaU7eG3Ha2TXZjM/ZT7n9zufGGMMAGkVabyx6w12l+9mVsKsVk3+w0KG8VX6V0yMmYhGaHBKzqZlo8NHE6gP7KpDUTiC4oZilmQt4csDXxJtjObWkbcyOmI0FqeFBFMCubXeNYD0ar3X98EhgwnQBXSlyCfE6c+s4pmLRrBwTGx3i6Kg0CYUy8ARpFWmceMvN7KtdBsF9QU8t/U5vjn4TdMbqk6t47ktz7GzbCcF9QV8nPYxFqeFAH0At468FX+tXCteIzRcN+w6jFq5KFy4XziPT33ca/mD4x8kJTClew60h5JRlcH1P1/PuqJ1FDUU8fqu13l397s4XA7y6/K58Zcb+TXvV0rMJXyY9iHFDcXMTpzdtH6iKZH+wf15c8+b/Gv9v3h08qMI5Op1Eb4R3Dv2Xny1vt11eKc0bsnNl+lf8uK2FylsKGRLyRau/+V60qvSeWz9Y1w64NImRU0t1Nw68laEJFAJ+TYV4hPCQxMfwqTv2SWpG2yy8llSa+1mSRQU2o5iGTiCA5UHvN4kAT5K+4gL+11ImG8YWTVZIPAKhPl0/6dMj5vO81uf56IBF6FT6wD4NfdXRoWNIj4gHoAJURP4Yt4XFNYXEuQTRGJAYo+NJuguMqszsbq8b6KLMxazaOgismuym3wutCotDreD/+36H5+e+ynzU+azq3wXZeYy3tglZ1HNrs0mxhjD53M/p8HRQJx/XJOFR6HrKTWX8nHaxwBoVJqma6igroCM6gxe3fEqC/otwEfjg1qoGR46nGCfYIaHD8ftdpMYmEiUX1Q3HkHbqGywA1BWpygDCr0HRRk4AoPW0KItQBeAVq2l3l6Pr8aXKwddia/Wl/y6fL7L/A6T3oSQBAODBmLUGnFLbkIMIVw/7HosTgvLspaRGJDI4JDBxPrHEuvfuumw3FLOztKdZNdmE2GIYEjoEJIDkzv7kLuVGlsNe8r3kFeXR4xfDAH6libgBFMCJeYS8uvyuX3U7cT5x1FmLiPIJ4jc2lz2Ve4jwT+BjKoMBocM5tKBl+Kv82dvxV4MGoMSPdBD0Kl1DAsZxvS46YT4hFBuKSfWP5ZqWzWLBi/CT+fHkqwlVFgquGLQFeTW5lLUUESkXyTFDcXoNXqC9EEtom8qLZXsrdjLobpDxPvHMyR0CEE+Qd10lFBtdgBQ1eDoNhkUFNqLogwcwZCQIUT6RjY5/gHcPeZuTDoTH+77kGe2PNPUPj5yPLMTZjMoZBCL0xczOnI0L21/qWn5uMhxBPsEMyRkCK/vfJ0npz3J4NDBre7X7XbzTcY3XuuPjxzPo5MfJc4/rtV1ejsOl4MP933I/3b9r6ltXvI8Lky9kMUHFze13Tf+Pm5bcRv1jnoADBoD/576b+5ZdU9TpMC5iecS5BPEy9tfblpvQeoCoo3RXXQ0Cscj2CeY+anzqbJV8dDah7h04KV8nv55k5+AQHD3mLvRqXW8uO1FLE4LAEatkeuHXc/VP13NU6c9xTlJ5zRt0+ww8+rOV/n8wOdNbVcNvorbR93ebSG7dVZZCai1KsqAQu9B8Rk4ggRTAm/OepPHJj/GHaPu4N3Z7zI1ZiqHag95PWgANhVvYnrcdJbnLCfAJ4C3d7/ttXxz8WZSAlL4Kv0rhoQOYVf5rqPud3/V/ibzdvPtt+Yt31fIrc3lrd1vebX9kPUD5ySfw+NTH+e2kbfxybmfsPrQ6iZFAORIg59zf/aymiQGJPJV+lde2/rm4DcUNxSj0DOoslaxIm8F6wvXY3fbMWqNXg6DEhJfpX+FQDQpAgD1jnpyanOI8oviP5v+Q5m5rGlZdk22lyIA8OG+D8mpzen04zkadTYnvjo1dVbn8TsrKPQQFMtAKyQGJJIYkOjVZnVZW4Q5gfxgmh43HX+tf6vL3ZKbCmsFgfrApvnu1rA6rS3mygHMzr4bE29xWnBJrhbtGqFhSMgQEkwJxJviKWloGX5ZZi7zigpojEU/ktZ+v3p7PXl1eQgE8aZ4/LR+J34QCm3G5rIhIVFhqUAgWh37CmsFDnfLN+pKayUB+gAyqjKarjO7y06FtaJFXwkJi8PSor2raLA5CfLVNTkSKij0BhTLQBuJNkYzNmKsV5tRa6S4objpjX5MxBiv5X5aP1ySi5kJM/mj4A8GB7c+RQCyRWJMuPf6vhpfkkxJHXQEPY9Y/1j6BfbzapsWPY1d5btY+MNCrlh2BZcvvZwzE85sse6M+BnsLtvd9L3aVk2MX4xXnxi/GOL9473a8uvyuf/3+7lkySVcvORiHln3CEX1LfMRKHQ8YYYwkgKSmBIzBQkJH41PU6RAIzMTZqJX6VusOyJsBBlVGZyXch4RvhFUW6t5dcerrC1Y2yI8N84/zisldVfTYHcR6Kulwd5S2VFQ6KkoykAb8df58/Ckhzkv+TyMWiMjw0Zy68hb+fzA57glN89seYarB1/NzISZGLVGRoSN4K7Rd2FxWoj0jeSWEbcwOmL0MfexaMgizk48G3+tPyNCR/DEtCf6dKriIJ8gnjrtKWYnzMaoNTIlegqXD76c57Y+h9Mtv1Xl1eVRZi7jvrH3EekXSbhvODePuBmdSsf5qedj0pnoF9iPiVETef705zkj7gyMWiNnxJ3BizNeJMw3zGufK/JWsCZ/TdP35TnLWVu4tkuP+1RFrVJzUf+LiDZGs2jIIn7L/Y37x91Pv8B+BOgDuKDfBfhqfPmj8A+uG3od4b7hRPpFcvvI29ldvpvLBl7GDcNvQKvWsqdiD2/veZvFGYu5avBVTIic0DTuL5z+AqGG0G47TrPNSYBBi0VRBhR6EaK1DG/dxdixY6UtW7Z0txjHxOFykF+fzyvbX+GXvF+8zNNXDb6KgvoCbhx+I0H6IFxuF07JiU6tI8oYhdVpRZKkViMWDlQeYOEPC5mTNIeJURPJq8vj/b3v89RpT7X6ZnyCiLZ27MqxsDltVNuqMelMrDi0ggd+f6BFn/vG3kdSQBIqoWJfxT4i/SI5Pe50CusL8dX6EuYbhsvtQiVU1NhqCNAHtFCkXG4X1yy/hu2l273aZ8TP4MUzXuzUYzyCNo8D9I7roj00Tp3ZXXIIXq21lq2lW/kh8wf2Ve4DZGfD+8bex6aSTUyLnsawsGGEGkJRq9QAfLD3A57e8nTTNsdFjmNoyFD+NOhP7Q0f7fCxeP6XdPYX17H2YDl7Hpt9zL4KXrRrLBQ6FsVnoJ1o1Vqi/KKwuqwt5qmTTEnYXXbu//1+5ifPR6PW8H3m94QZwrh04KV8m/EtpZZSrhl6DZOjJzfNVVudVmrttYQaQlmavZSl2Uubtnnkm21fRK/RE6GRb+DhhpZVBVMCU0gMSOSt3W8xJ3kOmTWZfJ/1Pbm1ucxNmUtBfQGPrXuMKlsVVw6+kumx01u1qKhVaiZFTWqhDIyPGN85B6bQKiqhIswQhiRJ7CjbwVu73qKwoZBpMdNIDkxmSdYSau216DV6fs75mTERYwjQBzQpAhaHpUXo4ObizRi1xlZDU7sai92FyUeDxaFYBhR6D8o0wQngo/FhYf+FXqbI6bHTCfIJ4rMDn1Frq6XaXs0zW54hvSqdtYVruXvl3QwNG8ru8t3cs+oethQffrvYVrqNO3+7k2uGXINBI1sNBII7R99Jv6B+LfbflwnyCWJ+yvym7/5af24YdgOv7HiFoWFDeXXnqyzJWkJ2TTav73qdpzY9xXt73mND8QYOVB3gobUPsfLQyqNu/5ykc+gf1L/p+4iwEZwWd1qnHpNC6xyoPMB1y69jTcEaDlYf5N2972LSmYgzxnHT8JvIqMrg7jF3s75wvZcCt6NsB2sL1jI9dnpTW6ghlJtH3NwjptXMdjmaAAkcrtYdWxUUehqKZeAEqLBU8PTmpzkn6Rz8NH6oVWp2le1id/ludCod02Kn8WP2j17rOCUntbZa9Go9NpeNzw58xmmxpyGEYEnWEuocdby7912uHHwlABGGCOalzGtSDk4VMqszyavL49aRt+J0O3G6nfxn839YkLoArVpLucW7nvyagjXcNPwm/ij8o6nt3b3vcmbCmZh0LdPWJgYk8sbMN8iqyUIlVCQFJBHsE9zpx6XQkv1V+7G77V5tS7KW8PeJf+eVHa9wdtLZmB1mlmQtAWBS9CQAlmbJ1rNpMdO4deStuNwuTHoTA4IHdPkxtIbF4SLYT4+PVoXF4UKrVt65FHo+naoMCCHigA+ACOTko29IktSlk7NtJas6i7TKNGwuG6E+oWhVWvoH9yfEENKir0alQS3UfLjvQ2YmzGRg8EAGhQwiwT+BEWEjsDqt+Gpa5r/XqDRN4VQB+gCEkKfIGguvRPlFNSkLOrWO9Kp0cutyMWgMDA4eTIx/TItt9lYK6gtIq0jD7DSTGpjKwOCBqIQKrVrL9tLtXm+CYYYwUoNSqbXVctOImxAIPtr3EXWOOjRCgxvvty9/rT8acfRTu9pWTZm5DIEg2CdYUQa6gGpbNXvL95Jfl0+wTzC+Wl+sTis3j7gZu8vOh/s+xO6246v1ZV3hOnJqc7zyDZh0JuwuO/sr99MvqB+3jryVcN/wprBTP42fXD+kB8w6WxwudBoVeq3aM2WgpBxX6Pm0WRkQQlwA/AcIR77kBCBJknSsqiFO4F5JkrYJIfyBrUKIXyRJ2ncyQnc0+yv3c+3ya5vyAJh0Jq4ddi3fZX7HgxMebFEYJUAfwO2jbufrg1/jklxeyYiuHnI1Gwo3MD91Pi9se6GpPVAfiFqlxul2ohEaLu5/cdOyc5LOYc2hNUyKnsTL219GIzTcO/ZeFi1f1ORVn2xK5r9n/bdPZCPMr8vnthW3kVmTCch5BV6f+ToToiYQ4xdDrDGW/Pr8pv7/N+H/eHT9o02lh41aIzcMv4Hntj7HxQMuZnPRZq/t3zTipqMWI9pXsY9rl1/blMQoUB/I27Pf9po6UOhYzA4zb+x8gw/TPgTgvnH38fL2l5se9GGGMK4bdh2v7nyVC/tdyIf7PmRi1ETKzGXo1Dq0Ki3zUubxR8Ef3LXyrqask0NChjAoeBBfZXyFRmiI8Y9hQtSEbjvORqwON3q1Cr1GhVXxG1DoJbTHMvAUME+SpLS2riBJUhFQ5Pm/TgiRBsQAPUoZWJa1zCshUK29lszqTHaW7ySjOqNF/gCAabHT0Kg03LHyDq/2T9M+5ZFJj+CSXDw7/Vl2lO0gwjeCoaFD2Vm2k5tH3MzUmKkMCRnStM6w0GE8cdoTXP/z9QBMjpnM0uylTYoAQFZtFjvLdvYJZWBX2a4mRQDkKZQXtr3AmzPfJLsmm3OSzsHpdlJuKWdI6BA2F29uUgRAzkiXVZ3Fg+MfZFf5LqbGTmVw6GDUQs0Z8WcwPHT4Uff9dcbXXtkMq23V/JLzi6IMdCLZNdlNisDQ0KFsLNrolWGwzFLWdL3k1eRx4/AbifKLwi25ya3N5YNzPiDSN9JLEQDYW7GX0+NOB+Rz6MVtL/LGzDcw6oxdenxHYnW40GpU6DQqxYlQodfQHmWgpD2KwJEIIRKBUcDGI9pvAG4AiI+Pb7liF5BVk+X1PcmUxJCQIQwMHojL3fJidrldFDUUed3QGrG77UT4RjAhWn5DmZU4q2nZkUpFna2OgoYC9Go9Rq2xKVteiE8I+yv3t9h2ubm8RVtH0lVjUWmtJDkgmTPjz0Qt1GTWZLKtZBsVlgp0ah1GrREftQ/+On8qLZUcqj/UYhv59fkU1BewuUS2Cph0Jv4+8e8tEkM1xy25W4w1tBz/nkBPuC46ikbly0ftw7mJ5/JD1g8t+lTbqr2ulUbckptDdYcoNhdTZilrsby5wlxYX4jFZcFIxyoD7R0Lq2eaQKdRYXUoDoQKvYPjerYIIS7wTBFsEUJ8LoS4rLHN035chBBGYDFwlyRJXjl5JUl6Q5KksZIkjQ0L654wurnJc5v+nxA5gamxU3l2y7M8tfkpXtj2glemu3p7PR/s+4CF3y8kvSodo9b7xpMamEq86fg3jOyabO5YeQcX/XARF3x/ActzlnPFwCsAT82DZp7SjQwNHXqih9gmumoshoUOY2LURN7b+x6v73qd4oZiHp/6OE9uepK7Vt3Fi9tfZG/FXg5WH2TVoVWcFX9Wi22cHns6u8sPj0udve64RYlUQsWC1AUt2psXvukp9ITroqOI848j0i+SO0ffyXeZ3zU5AjYnNTCV/Lp8rzaL08JX6V9x0Q8X8cTGJ5gRN8NruVqom8qFAyzot4BQn45PNtTesbA53ejUKnRqZZpAoffQFjfXeZ6PCTADs5q1zT3GegAIIbTIisDHkiR9feKidh4ToyZy39j7MOlMTIqe1OTMBLC7fDdv73mbBkcDIJsmn9v6HHa3nS/Tv+TWkbcyKHgQKqFiSvQUHpn0CFHGY9dcd7gdfLD3A7aUyOGFTreTV3e+yriocZwVfxbFDcWEG8L508A/oVfrCfcN5+nTnu50ZaCrqHfU88n+T5py0O8u3833md835Zl3S26+z/qeBFMCWbVZHKw6yP3j7sekM2HUGrlnzD1MjZ3KvJR5aFVaYo2xvDTjJQYFH79U8ZToKdw95m6MWiMmnYkHxj/AuIhxnXq8pzrRxmhePP1FPt7/MQeqDlBqLuXSAZdi0BgI0gdx75h7+TX31xYROAcqD/DPDf/E4rSwq3wXyYHJzE6cjUalIc4/jienPcnv+b+jU+m4YtAVXNT/oian3O6k0TKgV6YJFHoRx50mkCTpGgAhxBRJkrzytgohphxrXSFfmW8DaZIkPXcygnYmgT6BXDnkSmYlzmJ5zvKmdoPGwMiwkVRZqihpKCHeFO9VZa3WXsuzW55lZsJMbhlxCw7JwcjwkdhddrQq7VFvTJWWSn479FuL9tzaXJ6Y9gSV1kqMWiN+Wj+uHnI1OrWu1aiG3kpmdWaLtrWFa5mdOJu0ysMzUVk1WUT7RfNB2gf8uOBHZibMRJIkIvwicLgdPDDhAW4Ydj16tZ4gQ9siAoINwfx56J+ZkzQHIQThvi2THCl0PALR9Ob/W95vnBl/JreMuIVw33AK6gsQQqBX6am31WPUG3G4HOTV5Xlt441dbzA4ZDCfz/mcMN8wgnyCGBc5DrvLTphvGBpVz4iUtjvdaNUCnUaFTVEGFHoJ7bl6XgaOTK7fWltzpgBXAruFEDs8bQ9KkrSsHfvtMiL8IprCzGbEzWBA8ADWFa4j1BDK/sr9LM9eTqQx0msdp+SksKEQt+QmvzafX3N+5a09bzEqfBQX9LvAK2lQYX0hv+b+SnZNNgOCBrC+aL3XtmKNsfhofLzM3cezMvRGYowtQyQHBA0gr9b75h9rjOX3/N+5qN9F7Kvcx5fpX+KW3CxIXUBmdSZalQadWk92ZRqX9buIoTET2yxDO1PWKpwEW4s2kVuXT6A+kAmRE0gISGBD4QbMTjNzkuaQ4J/AobpDLMtZRr2zntHho/l0/6et+hBYHBb2VuxluGo4QT5BPVJJbpwm0KoVnwGF3sNxlQEhxCRgMhAmhLin2SIToD7WupIk/UGPiPxtO0NDh3Ju4rlEGaN4bedrTe1/FPzBDcNvoNRcyvTY6azOXw3IMe1nJ57N/b/fzw3Db+DLjC8J1AfyUdpH/Jr7K++f8z7RxmisTiv/3f5ffsj6AYHgr+P+yr7KfdTYagCYHjOd4WFH94LvS0T6RTIhcgIbi2VfUpPOxLVDr+WRdY809RkYPBCH24GExJSYKdy16q6mZZuKN/Hvqf/m0XWPMjF6IlanlT+vuoOPZ7xGv6iWkR8K3cfe0h3ctfpeAG4fdTs5tTlNVT4B1hWu46GJD/HNwW/k/hV7mZ0wmznJc/gxS7YG/ZL7CyBb6i4deCmPb3yccN9w3p39bo9U6mxON1qNrAwo0wQKvYW2WAZ0gNHT179Zey2wsDOE6k4SAxK5edh1XLfiZq92m8uG2WHm/X3v89z05xgVPgqr04pTcvLqjlexuqw0OBrYXLyZa4ddy9rCtRSbi8msziTaGE1BfUFTJjUJif9u/y+XDryUWGMs1bZqZifO7pE3ts5gX8U+THpTU/a4IJ8gPtj3AWclnEVqYCoWp4Vgn2CyarJ4YtoT/JT9U4ttLM1aytzkuXx98GtuGn4Tm4o3caA6XVEGehgZ1ZlNYaElDSV8d/A7r+UWp4VDdd7RIj/n/syM+Bn8euhXpkRPacpGadQaSauQE4MdqjtEVk1Wj7xmbE4XWpUKnVoo0wQKvYa2+AysBlYLId6TJCn3eP17CpIkkV6VTlZNFr4aXwYGD2zzjSNQ44tW1TJrmFqlRkiCKlsVizMWt7iJqYRKzrMoea/TuEwt1DglORTK7DTzzp53uHnEzXyV/hUL+rX0cu+raFQafsn9pemNb1zkOCRJ4qO0j7h5xM1eFpn5qfO9PMYb0al12Fw2VELVFHuuFsc0VCl0IQerD3Kw6iA0G5MqW1Wr15U4wnioFmr5WkL2JWksMT04eDDhfod9PBqvrZ6EJEnYnW50HsuA1alMEyj0DtoSWviDEOJ74GUhxPdHfrpAxhNiS8kWLlt6GfetuY/bfruNe1bdQ1F9UZvWDTTFcNvgRV5tJp0JtVBzZsKZ/JT9E1cPvtprub/WH61Ky4yEGU2+AP0Ckkn1iwXk+e/LB13utU6SKYkKawW3jbqtW+uvdzUpgSn4aw8bmbaVbOOSAZcAcgXHCN/DStt3B79jdsJsr/TCKqHi3ORzWZq9lHOTzmVt4VqC9EEMChrYdQehcFR2le3i8qWX89c1f0WtUhPtJ/vA/Jr7Kxf2v9Crb5A+qEX9jQX9FuByu1q0z06czbqCdQAMDx1OSkBKJx7FieF0SwghUKuEx2dAsQwo9A7aMk3wjOfvBUAk8JHn+2VASWcIdbLU2et4dsuzTaFrALvKd7GnYk+bHfJOTziT1/QBLC9YTZAhlJTAVMwOM8ERwfhofFCj5slpT7KuYB1hvmEkByShttYwLHIqy/JXcE7/PzHV5iK8ugAC4tGqtVwz5BqGhgxldf5qkgKSCPcNJ8QQwujwY/lg9j3KzGVcO+xaChsKqbZWMyRkCBlVGbw16y1W5q3kL2P/Ql5tHjm1OYyLGIfVYeXFM15kVf4qXJKL02NPJ71yPw9P+Dt2p5lkQwSnjzmNxIhTw+eiJ2Nz2Xh95+tNCbn+se4fPD7tcbJrssmsziTZlCxfN/l/kKD24YzEs3Ho/blt5G2kV6UzLnIcSQFJrMxbyT8n/5O0yjRq7bXMiJ+B3WlnVuIsRoSNYEr0lB7pPGh1uNB5ChNpNSosdkUZUOgdtHWaACHEs5IkNU/v9oMQYstRVutWzA5zC890kKsNthU/31CmJp/N1OSzoSIbbNXgFw4BMSBJUFsEkpM5saeDzg/KM+D1qYDE1KAkqCsGazVEjmraZqhvKJOiJ5ESlIJJZyLSL/Ioe+/bFDYU8sK2Fwg1hOKv8+eX3F8I1Adyaf+FTOh3MQgVxJ8Fas/p+fvzOPLWkzrmClBpCFv5HGeYYmD+q6BSgdsN1TlQdgC0RnmsfMPAXwkb7GosDotXRker28q9q+/l/tH38FSDgG1fgtAyJ3aMfH1Y6iFmAoNDBzetY3aYCTWEolPrmJU4yytEd0aCd+KhnobsPCjLqyQdUuhNtKe2pp8QIrnxixAiCfDreJFOnhBDCHNTWuZDah7m1yZcTtjzDbw7C944HT44D7LXwPaP4JWx8MIw+OwKWRHwDYXQ/uC0Qdl++Uan0UNwUtPmGovkXPj9hVyy5BJW5q1sNd1xX6cxOVC5pZzsmmwkJOYlnUvIutfg5dHw33Gw+klokNMvl8eM5IW4FOZseoRzNzzIf+L7UzzsQlkRMFfB2hfgnXMgcwW8Oxtemwxvnwk5a48hhUJnEOgTyPnJLa+9ZF0A5P4BEUOhNg9+uh/WvQyVmWBraOqXW5vLfWvu4/zvzufC7y/ko30fUW+vb7G9norN6Uavln0Z9BoVZsUyoNBLaI8ycDewSgixSgixGlgJ3NUpUp0kGpWGKwddybzkeaiEiiB9EE9MfcKrOFCbKNwG394I9aXy94qD8N1t4HaC3XMDy/oNfn9Otg4seA0Sp8ntQYlw+RcQJs9jV1mrePCPB0mvTgfk/Pz3rLrHq2DPqcKQ0CE8PuVxAvWBqISK81PO57Kg4ajWvShbXVx2WPM05Mrzw+tdNXyQvQSn5MQlufg872dWOzx1GvI3wYrHYPjFsPopqPE4dVbnwed/gqpe4/PaN5AkztNFcnG87Odh0pn4+6BrGG5ugEm3yddU3ga5r70elv0VinYAcibOj9M+bgrbtTgtPLXlKa+00z0dm8N12DKgVC1U6EW0OemQJEk/CSH6AY1eWvslSbJ1jlgnT5wpjkcnP8pNI29Cr9KfWAhSZbb8pu8fBUEJULofqnNBb4KhF4HOAHu/hbTvYcZD8lvPpZ+CrRa0BvA9nBWvxFzSIvOeU3KSX5d/ylXMM2gMnJd6HhOiJsiFnbQmdO+e690pKBEqcwBYkb+qxTaW5fzERQMvQXVok9yg0YGlyruTpQpq8uSxU+gaLFVE//IYfzMEsWjgQrR2C5F7f4HUmaDSwqGNkDhVvobK9kNllmwdSJxCtbXaKwNoIwcqD7Raz6An0phwCDzTBE5FGVDoHbQl6dAMSZJ+a6UoUYoQgp5abwDk8LN4/6MUDbJUQ/Zq2PaB/OAZOBf2fgO2OhhzDUQMBp9Ass96iJ8xs7k+h7MGz+B0q4PKoCi+i4gk11LKgvlPMbGuloDNb0HMGKgrkpWD+EnyZ+P/wBSNaexVBOgDmpIMNdKY8bCv4XLa2VW0ka8yv6PBaeai5PMYHTUeQ2Pa4IJtRGz/WH5Yj79Z/u1K9oDWlwNnPchSWzH7G3Yzd9/HzIyawoq8FV7bH2lKRrXlPYj0OA2qNPKnWRU7VBrogU5mfZLyDFkxri+BiGFoDywlrrYAptwlK9M5aymedCO75z9Dvr2a9YXrSEiYzwJdJIP9Zadeo87IwOCBrCtc57Xp4xWg6knYPGGF0OhAqIQWKvQO2mIZmA78hlyY6EgkoMcqA8dk33fwwx2Hv+/8FKbeDb/9C/Z+DRe9T6mPkbvK1pBVmwPAxpIt+Ex+jCdW391Ubnht0XoeGnkHl+QXQFW2rFCA7FsQmAADzoaN/yP64M88PO9J/rr2/3BL8g3iqsFXkRqY2pVH3WXsKdrENStvxyXJb0YrCn7n5SlPcHrqXCjeDe/NAYf8G5LxM1z+JRxYSu6E67g++0uqbPJb/vriDdw49FomhI9lY6nsrxphCGOeKhCW3g3DLoKRf5IfRBNvgXUvHRZi1r8gpJ1+Igrtp64IPr9CftMPGwgzH4PctfJ4rHwcHBZsceP5uHgdwsfEu3vfBWA9sERr5MOZb5IK+Gh8uHXkrews29lUGGxcxDiGh/aeKBGrw4XWYxnQK9MECr2ItkQTPOL5e03ni9NF1JfBmqe82+wN4LSCSg1uF+z6nMxRC5sUAZCToxQ0FDUpAo28fuATzhzzd0I/8o6hpjoXfAI9/x/ijAYzn8/5nEP1hwjxCaFfUD+Muo6tvd5TWFWwpkkRaOS99M+ZHH8GuoKthxWBRn7+P7h6KQcqd1OV7W3ufz/tYz4Z8wDFiWfjri0ktb6amF//JS/c8xVc9jkExkP4ILh6iTw9YIqRv2taJrlR6GDKDsiKAMDsf8PKf8OEG+XvDjnEML//WVg1TpYf/NZr1XpHPftrc0gNkytyDg8bzmdzPiOrJguDxkD/oP49MoTwaDSfJlCUAYXeRJt9BoQQmcAG4Hfgd0mS9naaVJ2F0w6le8Fa24bOh/OiaVQaLh94OX5aP0INoSwasoiP0z5uymMgEKDzbX0zAbHgFwYNZWgtVQw0JTIw5NRMjiP/ohLNy1VYo0ZwcMSFFKpVhNkr8NWZWl3X1y0xzWIGswXcEjSvCGmvh1VPwIVvQ9K0zj0IhcNU5kDpPtnhsxGNDzgtsP5VmPs8nP432Sk0rD9U72u1kqcQAqvTSkZVBoUNhUT4RjAxaiK+2qNcUz0Yq8OF1jNNoNOoldoECr2G9lQtHAxMAKYBTwshBgC7JEnqPXl0930L39wgv62Pu072WG9E7w9qvWwVEAKGX0KKtY5UUxKzk89lccZiihuKAQj3DeemETfx8vaXAbh58NWEbn4HhiyAPYsPbzMoCUr2wbR7YdWTsmPhnsUw8nLvh1kf5PSY03gv/cum9MsA1wy4FJ3OKPsH6PxwGyP4YdT5/CPtLbnDPrhxyDVMiZrC2qLDYYHXDbqS6I1vQZ5nLtkYAdPvhxX/gGGXQOEOeUxD+uaUS4+kZC98eL4caTPmGjms1lIFDRXytVVfKkd5eKI74vrPxpA8igWpC3h7z9tNm/HX+jM4eDBfZ3zNE5ueaGq/c/SdXDnoSvQafVcf2Ulh85QvhkbLgOIzoNA7aI8y4AIcnr9uoNTz6R1UZsPSe+S3FEsVFGyDGX+Hgq3yQ7vfLNnxb/ilMPpKCB9MWN56Xhr3IF8WrW1SBABKzaVUWiq4MGU+U8NGM66+BnZ9DqOulOepD66AkBTwC5UVjoih8lvS0nvkt6j4ifLyPsyw6Im8O+NlvstaQr3TwgVJcxgVNUFeGDkUFi0ltyabJ7c87rXeG3vf460pTzDEEMYBaxnnRk9lgs2FyGvmVFZfIocOzvq37KcR2l/292go68IjPMXZ/vHhkNtt78O0eyBusuzHMXi+7LxZc7h2hy59OX8aPI89vkb+OvYvbCzaRGJAIvOS5+HCxdNbnvba/MvbX2ZazDQGBA/oyqM6aWzNMhAqoYUKvYn2KAO1wG7gOeBNSZLans6vJ2CtkSMFGslcIUcTLFoKxiiw18lKgEoHjgb5wTJwDnFAVtp7LTaXUZXB29Oekftu/avcuOMjmP6AbAFI+0F+aIHsaV2w5XDom7W6M4+0R6BSaxgZO5WRsVPlhqpcqMgE/0gwRUP0KGqxYXfbvdaTkHDY6rhdhMpKmggAPw3MewlW/wdqC+SOZQfAJwA2vwkzHobf/gHn/beLj/IUxFwpXxs6g5xUK2I4TL1LngoTKlj4DtjqZSdRra+Xb0jE9k+J0PiAvZ6rFr4jT6Eh16ZwNo8CAdySu8mJtDdhdbq9HAiVaQKF3kJ7lIHLgKnALcB1Qoh1wBpJklYce7UegilG9iyvyDjcZoyUb267v5Krq6WcIWe+K9wuO6TNehwGzGFm1BRWF3pns5sTMkKOHDi0AUZcCuk/ylYHtUa2NjQnZQbkysWLCEwEU1znHmtPI/1nOeNcZRZEDIFzn4eKdCIr04nwjaDEfLjEhUFjIFZngg1/l988E6fJv1/jQ3/r+3BoPQw4Fza+Lk8NVB6UV1amCTqXQ5vh+zugbJ8cNXDxh1CVB1U5sOROOWNk8hmQOAV2fAJnPSZfT2bPe0PceNj8lhwO6hPQtNkoYxQhPiFUWA+/X/hr/YkxxnTxAZ48Nq9oAjVWhwtJklr1lVBQ6Em0OQOhJEnfSZL0V+BGYBmwCFjSSXJ1PMYwuPAt2WQPcuzz+a/Ad7eCIRBCkuDXh2VFAGQz9NfXQf4mJvsncc2Ay9CoNGiEhisS5zC1+CD8/ACEJMOBZfIctloLB5bDWY/KSVUAUs+CuAly5rWwwXDRu6dWzvyinbD4WlkRAHmuuWAzfH8bEZve5rnUy0j2l5MCRflF8fKEh0n49o7DJuic3yFrlazIfX8bTLwJhl8iWxf8o2HctZDxq1ynIGpE9xzjqUB1Pnx2mawIAJSnyxYCe60cCeJJHU3WSjmxkDFc9hkYe408ZTDiMnlMwwbBuc/IPjoeovyieOGMF0g0JQJyhc+XZrxEnH/vU5qtDhcaj8+AWiVXL7QpZYwVegHtiSZYDIwAMoE1wFXAxk6Sq3OIHimHntUXywpA9u/ywzx/i+zUV3bAu7/TBpVZhKV9zx0D53Fh/EXyZtKWoS32pEh1u2RlYPKdcP5r4LCBkGD+KxA2QDaF2uqg30z5Bunbe8KkOoSKTHnapBGtQY5LB7A3MHzpA7w/8x+UqyQCy9IJLdjXcu4/ezWc9hf5r6VSDmOryYfLPgNLBQycB4G978HRq6jO9R6XqBGQvxX8WjmfD66AqffIaYdNsbIFQecLtcVw9pPg0zJiZGT4SN4/+30qrZUE+QT1qnDC5lia+QwA+GjVWOwufLTqbpRKQeH4tGea4AlguyRJrU6CCSFmSpL0S8eI1Yn4BskfAEMwDJh7eG5fZ5TD1JpjCITRi9BIThKWv+rtdwCg1skZDEt2g+SS57UlN0y+Awaf59munzxXfipypPLjtHk/DM7+D4FqLYEHV8jJns74v5bbCIiVc0OAbHEZdpG8HWOo/FHofAxBh3NwAAgtRA0FVSve/oEJUFco/6/3l8cu8xfZSfcY5vJgQzDBjRkqeylWx+EMhAA+GjVmh4ugbpRJQaEttGeaYMvRFAEP/+kAebqWoCTY/aXs9X9oixxd0JwRl4K1Ti6Pu/5VudBKc4YsgPzNcga86kPy3Kjklrc34pIuO4weTcQwGH/T4e+SW/b+jxkr/54Hf5WnamJGy+GB5emQNP1wf5UaJtwkKwoD58qWhl/+DlG9JytdnyAkVfbZAPmBPnS+PF6HNsp+Ao2oNDD+Ojkj5NhrZX8clZAdar+/Xc422Yex2F3eyoBORYPNeYw1FBR6Bu2xDByP3uchU3tITkK05qBs1t/3Ayx8V7YU+ATKUQDV2bDtQzlXwJa35YJEGh/Z+VDnJ/sJqLTynHVlJiSdJjvJBScfd/enBH7BcthfynSoLZLfGqNHQdxE2X/gY0/WxtX/kbPWCTUMuRCGXgAuBwQlQ10xnPOUPDWw0hOKuP1jufiN4pjVNWh0sn9G/ETZerbjU/DNhe0fwPn/g/6z5esgIE6eTpjxdzmaYMvbEN3Ml2PVk5Bypmxx64NYHC78fQ7fVn00aqWMsUKvoCOVAakDt9Xx1BQcTpkaNhACYuS3VJBNzuXpsP97SPvOe73p98vFb1Qa+eFV4km8OOg8uORD774Rgzr3GHorpkgweaoSVmTKPho+/nKmuun3gdsNm96ANc/Ifea/Aj/cKedt2P8D+ATJCaOa4+yxBTN7Ny4nlB+QIwSMEfK1ovekzNb7y8qA2yUrY43XT3W2/JA/kun3y3+lZrcGt/Pwen0Qi+MIy4BWrVgGFHoFHakM9FzK0uHTS+U3d4DgFNn5LDhFnu+vyoHcddD/HNkZsJGAODk/wYQb5TC25oz9c1dJ33co2AofLpAVqapsyPlDbvcJgDMehOUPyFEDLoccsx42ALZ/CBe8JStpzR8q469TrAKdwYGl8NU1h30Dpv9N9n/R+x3uU7hdziIpuWQl2VwhT7lVZR/u028mHNok+xo0f/hPu9ertHdfw3qEMmDQqalXlAGFXkBHKgM5HbitjmXft4cVAZD/3/cd9D9bnu+vOQSlaTByLkSNhIO/QOQIuYyxSi3PYUePho1vAC4Yf6P8hqTQduwN8NvjcmRBYJz8kG/EWgNpS2D2UxAUD7s+hfNfh/1L5WRDCVPhym/l39/lkMMLEyZ326H0Wapy5TwC7mZm7dVPyg/22LHyd3sDrPinPI6jroS5L8h+AJPvgIqDkL9JdhRUa2Vr3LR7YfcXcr6Icdd6+xf0Qcz2I6IJNCrqrYoyoNDzOa4yIIS44FjLJUn62vO3RT8hxDvAXKBUkqShJyrkCeN0gLkc8ta3XJa3Tr5BrXlazq0+6krI+EUuPbzgDdkMHT9R9l6318s+AAvfkv0FVEqYUKs0VMi/lV+4nKEOoLZQngaQXBA/AUIHyN+PpHQvzH1RNknHjJJN1EMvlJM4AQREQeJpgKT8/p2Fpbr17JjVebJ/jD4ALOVyeGjsWDlfRPUhOaGX2wkIuOoHeexdzsNjlzhVHnN13x836xFhhIplQKG30JZognnH+Mw9zrrvAWefhHwnTlWO/KAv3C57rh/JoPPkGOlZ/5Tj13/wJLo5/QHZVPrbP+SiQr/8Hba+By8Oh8+uOOx3oHAYt1tODPTOLHhphJysqTQNNr8N75wt/56fXyGXti3cDsFJLbeRNB1WPS7Xh3hrpryuvcG7j0qlKAKdiSlKzpDZHJVGNv+/OQPWPi9XKnQ74I9n5TwCA86BnDWypUBngGV/gYqsw4oAyNM5p4AiAGA+cppAq6bO6uhGiRQU2sZxlQFJkq45xueYE+eSJK0BKjtM2rbidskOaeED5QdTfYkcBihU8mfUlbIpE2STZ+MDPnet/OB32uS2Xx+R31TL98vz11m/weLr5RTGCocp2w8fXySbiSVJ9rvIWi0XZhqyQC4vXLxL7ntoPRz8DabeK+e2BzlDY8RgOXxw1b/ldZbeI6d6Vug6jOGw8G15/h/kuf0Zf5dTQDvM8jVVuk/2BZAkOavkr4/CWf+Qr7maQ3J9jtVPnrIOnha7C72Xz4CGWotiGVDo+bQ5zwCAEGKOEOI+IcTDjZ/OEuykqC+VvZ0tVfLb5bb35QI30+6VP6OvlqMJKrNbJhkq2SOnuoXDRVYO/AjJp8v/l+6VzaYKh6k46F3TPihRnjsGee64sUBTI7s/l60yk26Tp2SMkbLCAHJftVb+P6Pn57Dqc8SOhWt/hhvXwBkPwfqX5eyDjWT8DLHjDn932WWrQMZPsPMzuW3PV1BXwqmI1eFCrzlsBfHTqak024+xhoJCz6DNyoAQ4nXgEuB25JwCFwEJJyuAEOIGIcQWIcSWsrIOKkGr9YWgBDmjYCOHNsnTBuv/e/hh00paVLQGbwcqtVaeE228uWl8Doda9TFOeCyaFZ0B5Ae6f5T8f+Nv3ZxGBeH3Z2UT9P4fvCMFGtc5hXM1dMp10VaM4XK64drCwzUHGgmIldN5N8dcIZcEb8Q/Wr6O+gjtGQuLw4WP9vBt1ajXUGNRlAGFnk97LAOTJUm6CqiSJOkxYBLQ/2QFkCTpDUmSxkqSNDYsLOxkNydjCICZ/5RzBwxe4L3s9L8dLlYUPkieMmjOhJvkaoQgV8Yr3AGjrjgccjjzH3IinD7ICY9FxBAYNO/wd0uV7DXuHy07ZA67yLv/uOvkjHShA+Q49uahZ8Muktfxj4bUM0/ugHoxnXJdtJf+s2VHzkb0/jB04eFcGyBPtzVP0S1UMOcZuTBYH6GtYyFJUgvLgNFHQ5VZ8RlQ6Pm0J7TQ4vlrFkJEAxVAVMeL1EEkTpOzCFqrYdBcTyx0ouxM2Ojc5BMgVxgccoH8thMYL89jByfL3tNaP3ne1GGDeS/Kc6mRw2VHNoXD+IXCuc/K0y8NZXLq2shhcM1Sua69xiBnFDRXyXHnToscuhk1XP7NFy2TTdGGIDkDodMir38KWwZ6BHHj4fIvPP4eQnYwtFTDZV+AuUxWFALi5MJT4UPkKYPIofLYnYJYHW40ahVq1eH8F/4+Wirru9AyYDfLlT7jJvTZLI8KnUN7lIElQohA4GlgG3LGwbeOtYIQ4lPgdCBUCJEPPCJJ0tsnJmo7Uam806AeDb9QSJ3h3dZ8TlShbfhHgP9M77bg5LY90BOnAFM6RSyFkyR6pPw5FuEDu0KSHk+9zYnvEdUJTT4aKhu6SBlwu+GjC+QXH3sDXPOjPF2qoNAG2qMMPCVJkg1YLIRYAvgA1mOtIEnSZScjnIKCgkJvwWx3YtAdqQxoqbM5cbjcaNWdbFFM+0623Jz7jByi++llcMPKw1E7CgrHoD1nZ1PmHkmSbJIk1TRvU1BQUDiVqbM6MRxhGVCpBIG+WsrquiDUcvNbsu+OUMGg+fI06OreV0xWoXs4rjIghIgUQowBDEKIUUKI0Z7P6YBvZwuooKCg0Buot7W0DACEGfUUVltaWaMDaSiXnZ3jJsjfhZBrqmx+Ww6hVlA4Dm2ZJpgNLAJigeeatdcCD3aCTAoKCgq9jvpWLAMAYf56civMjE3sxAJNGb/IpcGbTwn4hsgRUWuehvNf7bx9K/QJjqsMSJL0PvC+EOJCSZIWd4FMCgoKCr2OOpsD31YsA5EmHzJK61pZowPJ+FnODXEkA+fCNzfKodZ+IZ0rg0Kvpj0+A2uFEG8LIX4EEEIMFkJc20lyKSgoKPQqai2tTxMkhPix81BN5+3Y7Zbrq0SParnMxwRx4+SskAoKx6A9ysC7wHLAk6uXdOCujhaooymoMvPt9nyeWX6AX/aVUNlwauZMP5WwO91syank5d8yeH9dNgeKa7tbpC6jpNbK0t1FPP3TfpbtLqKk9pgBPwodSK3F4VWxsJGBkf7syq/uPCfCsjQ5l0fzBFHNSZwGu7/snH0r9BnaE1oYKknSF0KIBwAkSXIKIVzHW6k7qai38ZevdrE+s6Kp7bYzUrjzrP6dH+aj0G1syKrg6nc3NWU4DjBo+eLGiQyIbCX9dB+i3urgyR/38832gqa2hWNiePS8oRj17bnUFU6Eaouj1d/ZT69hzvAoTn9mJe8uGs/4pA72HchaJSfwOhpRI+XU3w3lcl4VBYVWaM8TsUEIEYKcbAghxESgE21fJ096SZ2XIgDw+uoscivM3SSRQmdjtjt5aUW6V6mDGouDjdl9v9JkZlmDlyIA8NXWArLLG46yhkJHUtlgx+8oSteCUbFcNj6ed/7oBM/+jF/lzKhHQ62V/Qkyf+v4fSv0GdqjDNwDfA8kCyHWAh8gFy3qsVgd7hZtTrdEg00pKdqXqLM62JFXxcasCirr7a3mgq+19P388DZn64Y6i10537uCKrMd/2NYYIbHBLA5p4OVUocF8jfKb//HInIYHFxx1MXV1mruWnkXoz8czeyvZvPmrjexOpUpplOJ9igD+4BvgM1ACfAmst9Aj6VfuJFQo86rbUJSMD/sLCCnQnlb6gsU11h48JvdnP/qOi55YwO3fLKVqyYlevURAiYm931P6qRQP1LDvStqpoQZqai3Y1YUgk6nusGB0efoykCoUY/N6aa8vgN9B7JWy7VAjldJNWqEXLOgFRxuBzf/ejMqoeKFM17g+uHXs75oPed9ex6/57e+jkLfoz3KwAfAQODfwMvIFQs/7AyhOorYYF/e//N4zhkaSWyQgYVjYhmXFMxbf+Tw675Ts956X2NrbhU/7Cxq+r4rvxaH082/LxhKSpgfYxMC+eCa8QyPDew+IbuIMH8fnr9kBHOHRxEbZGDu8CgWjIrhtk+3k15c393i9XkqzXZMPq2U7PYghCAuyEBWWQe+iOz56nCioWNhigWnFapyWyz6YO8HqFVqLh1wKQaNgQRTAjePuJnLBl7GP9b/g1t+vYX9lfs7TmaFHkl7vIqGSpI0uNn3lUKIfR0tUEczJDqA0fFBnNYvlCqznfSSBkw+GlYdKOO6aclklNThp1MTHSQnU3S43B5zn7bVMCGFnsW+Qu9IgUvGxhLkp2V6/1DmDotGoxb46uTTvPnYWh0uhAAkqLLYiQn0Rac5vm5cZ3Vgd7oJMfasfO81ZjsuCTRCMCExmDMGhLE6vYy3/shi/shozPa+P03S3VQ2HFsZAAg3+ZBT3tAxToTmSkj/Cea/dvy+QkDEMMhd61W8qNZeyzt73uFv4/+GEMJrlaGhQ3ls8mOsPLSSm365iVBDKKfFnsbI8JGMDh+NUXcca4RCr6I9ysA2IcRESZI2AAghJgBbOkesjmN3fjVldVYsDi2Lt+UD8OepSYyIDeCpn/azeFs+oUY9t56RyuAof978PZvle4sZHhPAXTP7nxJvlL2ZoTEBAPjqVLx46Si+2JzPv5ftZ/neIG6cnsyYBPmmm1VWz9t/ZFNjtjM4OoDPNh9CrRJcOi6OLTlVhBh1XDExnqExga3ux+Fysz6zgqeW76e8zs5VkxK4cHQMEQGGrjrUVqm3OVm5v4QvNh/iqsmJvPNHNgdLG5jWP5Szh0YyNCaAD9bnsjOvmrtn9ueMgeFHdXJTOHFsThdWhwtf/bFfIEKNOnIrO8gysPo/cthgW0sVhw2AnD9g5OVNTV/s/4JhocOI9ItsdRWtWsusxFmcGX8m6VXp7K/cz2s7XiOrJou5KXO5Z8w9+Gn9OuBgFLqb9kwTjAHWCSFyhBA5yEWKxgkhdgshdnWKdCdJbkUD17y3GZNBx3O/pJNbYSa3wswLv2ZQVGPljTVZlNTa2FtYy22fbGN/cR0fb8yjvN7ObwfKuPLtTeQqvgU9mjGJQVw6Lo6H5w7hn0vS+CWthIoGOz/vK+Huz3eSXVZPndXB/327my+35NM/0sRTyw+QV2kmu7yBJ37cz+iEQL7Ycojnfk6nxtx6udndBTUsencTewpqKa618tTyA3zlUS67k83Zldz+6Q4uHhfHnZ/tYH1WJWX1NpbtLuJQpYV/L9tPfpWFzPIGbvt0O1tyq7pb5D5JRb2dAF8tqiPero8kzN+HvJONZrJUw/KHYN93MPJPbV8vYqhsGfDgdDv5ZP8nnJVw1nFXVavUDAoZxIJ+C7hn7D38e9q/Kaov4oplV1BlVc6pvkB7lIGzgSRguueT5GmbC8zreNFOnpzyBmICDa168C7fW9L0VgngliCrrAFfnQpfnRoh5JC0Dp3fU+hwwv19eHjeYIJ8deRVet9kS+usZJU1cKjKzPrMSkbEBbQINQXYllfNgEh/fjtQRtZRwvB259fglrzb3luXS1ld93pc/7CrEJBzDOg1Kny08iU9JiGI5Xtb+sX8uLuoRZvCyVNaZyPYV3fcfmFGHfknU7SoaBe8Mh7KD8A5T8mVCdtKUII8tVBXDMDq/NUE+QSRYEo4zootMelMLBqyiNTAVO5aeRdOt+Kg2ttps71QkqSWnic9gKIaCzvyqskqbyA6wAe9VkVUgIHs8gZK62xcMyWJ9JI6VqeXea0X7Nfy4WEyaLnl9H5UNtgIMGgprLG2WnhEoX2U1lrZcaianIoG+oX7MzIukCC/4984m5NWVMv2vCokCUbGBzIk+vBN0Fenwb+ZF3d8sC+XjY+jyuygoMZCgK8Wk4+GWouT5LCWJs1gPx1pRbX46dRND9Mj8W/FSzzUT4dO073nR/9wI69dMRokuH1GKlanG4NWzabsSny0LUNrI0w+3SBl36e01kqg77H9BUAuWlRQdYLKgMMKn18Bo6+CpOntX1+oIHKoPFUwbCGf7/+caTHTTkwWZIfIhf0X8vzW53l3z7tcP/z6E96WQvfTqycPKxvsPPjNblbuP/ygf+7iEdz2yXYKPNq3EPDMwhEE+Wqb4s8NWjXzRkTx7Y7DCVqiA3wI9NXy0Ld7mtompYQQE9S9c8K9nVqLg8eXpfHdjsKmtpumJ3P3zP7o2/gg3VNQwyX/W0+DXY6j12tUfH7jJEbGBTb1SQ7z4/xR0fyws4hFkxN5fFkaLs+rfKhRx5MXDueWj7dx0dhYfttfis0pPyj9dGrOGRrJ55sPcd/sAQyI8G9VhlHxQUQF+FBUI1sChID7zh5IgOH4D4DOZHhsIBuyK/hqSz6FzWR74JxBxAT68Ht6OXaXfKz+eg2zhhwlZa3CSVFSayWoDZaBYD89VWY7dqe7TQ6rXmx7H0zRJ6YINBI+BLJWkZ84kb0Ve7l6yNUnvi1AJVRcPeRq/rnhn5ydeDZxpriT2p5C99GrlYH0kjovRUAlIK/S3KQIAEgSvLY6k0fPG8L+Irly2ITkYDJL6nhv0Th2HKrGZNAyLCaAq9/Z5LX99ZkVFFRbiAv27ZoD6oMcLKv3UgQA3liTxYJRMW1OD/zDrsImRQDA5nTz6aY8L2WgrN7OgAh/5l4ZzeurMpsUAYDyejv5VRY+unY82/OqeO7iEWSVNWBxuDDo1KzPrOCz6yeQEmZEpWr9Bp0U6sfH101ge141NRYHI+ICGHYUZ8Ou5Ne0Evx9tE2KAMjn/OJth3h47mBe+dMotuRUERngw+SUkD6fkrm7KKqxtkkxVKsEIX46CqstJIa2w/FOkmDj6zD+hpOQEjnfwOr/8GVMChOjJqJTt89C1xqhhlBmJcziqc1P8fKZL5/09hS6h16tDLjdEnecmYpKCOosDvwNWhKCfZk5OIJfPHkE4oN9OW9ENJIkMTTGRH6VhUCDhtRIf7QqFaf3D8PmkhOBPL5gGM//mu6VrvhUyFzXmVjsLbPiuSWwONpW1qKsztqqWbW4xsL+4hpyys2EGvW43BKhfjqi/OU3rxZyOBwE++kw+mipsTjw06uRgO151ejUggGRyWSWNVBptpMcakStEuRUNFDRYCPC34f4ED+Sw4wkh/WMcKr8KjNmm5OKehsmg5aLRsdy5uBwAIx6DbU2B346DYEGDaH+ehJDfIkJVKxcnUVBlYXIgLZNwYSbfMitNLdPGSjYCi6n/GZ/MgQmIDksbN73KZdPefjkttWMWQmzeGjtQ2wt2cqYiDEdtl2FrqPXKgMbsyp48Jvd5FSYuXZqEmnFdazzOIdNSg7hz1MSWZNRzpxhUbyy8iA2p5tAXy0PnjuIb3cU8v76XCQJEkN8uWJiAv9amkaAQcvj5w/l8aVpFNXKmv7B0nrig2sZGKW8UZ0ISaF+hPvrKW1WsW1ItIn44OPfCLfmVnHLR1u5/5yBLNl12PFNCPjzlCSue38ro+ODcLkllnoc4wZE+HPPzP7c+sn2pv7+eg0DIwO4/K2NVJtlR7ubpqewbHcR4SY9V01KZt7Lf5BfbUWjEjwybzAmg5b7F+/C6nBj8tHw8mWjmD4gvAN/mRNn7cFybvl4G9dNSWTB6FhUAsrrbdz88TYkSf7Nr5qUQFmtjUCDlu05FTy+NI1zh0Xx0JxBRCtKQYdTUG3xckg+FhEmPXkVDUBY23ew+ytInCqf/CeDEBQERnOe00WEX8dNGWnVWuanzueZzc/wyZxPWuQsUOj59MrSfZX1Nl5dlUlOhZkgj9POumZe4uuzKtBr1Fw6Lo6Xf8tomh+uNjt4evkBXG6aCtnkVJhZk17G+KRgaiwOHluyj7+ePYCJycHceWY/Xl2VyXM/p2N1KN6yJ0J0oIF3F41j1uAIgv10XDg6hucvGUnwcRwIK+pt/OXLHaiEwOFyc/dZ/UgI8SU2yMAzC4fz+upMimqspIT5NSkCAAdK6vhtfyn3nzOASJMP/cON/GfhcB7+bg/VHp8Rm9PNy79lMG9ENGsPVpBWVEuNRR5fp1uivN7OX77c2VTbotbq5I7PdnCosvsLXBVWW7jzs+2EGnXU2pxklsmOsh9tyGs6p7PLG1idXsbG7Eo2ZFdy5eQkAJbtLuKPg+XdKH3fpbDaQmgbE1FF+BtIL2lHRkhJgn3fQsLkExOuGTaXjaXOcmbVd3yU1MSoidQ56liRd/QaCAo9l16pDBTXWtniCRdMDjO2yEIHsLugmhCjrkU4WFmdjYAjvH6351UzyPPmX1Znw6BVoxKCfy7dR73NydrM8laL3yi0jSExAbx02Uh+unMaT1wwjP5HcdJrTkW9nexyM+OTg/lpTzFv/Z7NqLggJiQFY9Cp2ZxTRbCvjoLqlqF9G7Mrya+0cFr/MG6cnkxprZWSWu988G6JJse6jVmVxIfIfiE6tQp/Hw0Ol/eJU2NxdF49+nZQXm+jvN7OwEgTIX56dh6q4lBly2mU7XnVBPvpCTHqKas7PG3SWmilwsnhcLkpq7e1qINyNOKCDewvbnnPOiqF20ClhcD2hwAeyQ+ZS6gPTcG/oRJ9dcfmyVAJFRekXsBzW5/D4Vbul72NXqcMuNwSkgQjPM5jOeUNDIhs+XAZEx+EXqNqYVUL8dO1qFo4NDaAjJK6w8vtTtZlVjS9aY1LDCbQcPKONqcyPloN4SafNofiBfnpiA0ysCW3ktHxQdTZnHy7o4DF2wpIK6pjeFwAVWY7Ua3M0w6LCWB7XjXf7SjA30eL3elucaMWQo5KABgcbeJQlZlBUf7cO6s/KpVArfI+cfz1mjbf7DuTED89Qb5aMkrrqKi3MiDCRFRg679BldlGVYONUP/Dco9P7IA0uApeFFZbCPbToVG37XaaGOLHvqJa3Ee+qRyNvd9C/MSTniLYV5nG6vzVnBZ3JjVxYwnf8+1Jba81hoYOJUgfxGf7P+vwbSt0Lr1OGdhTUMOC19by5ylJRJj0VDTYMWjVDI85PKc/Mi6QlHAj//0tk5tOS2m6sfvq1Pxl9gCvkq4RJj1nDQpnfVYFvjo1/5g/xMuBMMKk5y+zByh1CrqYMH89zywcQb3VydjEIEbHBzYtW32glHvOGkCQr+yVffqAw3Ov8cG+DI8NILOsnofmDOLfS9N4ZVUmN01PwdczhmqV4MbTkvlpTzETk4I5c1A4KiE4f2QMT/y4n4835HLrGSlo1fJ546NV8dwlI4gP6f60qzFBBp69eCSHKi0E+ukZFmsCt8T5I6Ob+kSafDhrUDhjE4IZEx/Ed55MidNSQ5jWL7S7RO+zZJU1EN2OtNQmg5ZgPx17W7FotkCSYM/Xsr/ASbCnfC+v7XiVuclz8dcZqU6cQmj6r2jry46/cjsQQnDRgIv4387/UW5RpqR6E0KS2qiddgFjx46Vtmw5drmDjzbk8tC3ewj01fDQnMG43BK+OjX9w/3JrWxACEGIn5brP9hGRYOdfuFG5gyPIjbQQESAD7VmOw63RG6FGZckMTUlFKfbTaXZQXywL0OjTZgdbjLL6jHbnCSG+BHVdxyu2vxq0Zax6ApKay08+sM+RsQGEuavR6cWhPr7UFFvIzrQwKFKM+EmPUjgkiR8tGoKqi1E+vtwqMrMvV/KmbLDjHouGhuLVi2YkhqKRiVwS5ASbmxSKl5fnckH6+XcWkmhfpw3IhofrYozB0WQGmZEpeowp6h2bejIsSioMvPZpkOoVDA6Pgi1CgQCu8uNxePw6HJLmO0uogMNWOxOtBoVKWFGAtsQC3+KcVJjAfC/1ZnszK/myomJbd7OZ5vzCDXq+fvcwcfumLsOvr0F5r10QpYBs9PM4vSv2VKymbnJc4nzP5wHICT9Z7TmKjLOfVxOSNSBfJX+FVaXlRdOf6E9zoSK12E30unRBEKIs4EXATXwliRJT57M9hrf0KvNTv7iudGnhPux+KbJDIgy8fW2fJbtLsboo6GiwU5GaT0v/JoBwJ1n9uOl3zJorv9MSg5hcqq3V69Rr2KEUqCoRxBuMhBo0PLEj94lVM8dFokKwfJ9xfx052mkhB8O+RvjmVot23V4jr/M43QaafLhyomJhPp7O3tFBxq8HMCyyxt4cUUGZw4M57qpyR2pCJw0GrWKjzflUdlgR60S3HpGCi+tOOjVJyXMj5Fxgfx5ShITUxRrQGeyK7+GuKD25SKZOSiC//t2DyFGHWPigxifFNz6Q3PTm5B6ZrsVAQmJtQXr+DL9C5IDUlg0ZBEGjfdLTUXqDOI2vEn8H/8lb+ptHaoQzE+Zz782/ovFGYtZ2H9hh21XofPo1GkCIYQaeAU4BxgMXCaEOI4qfGxGxgUSdoTX7n2zBja98YyMC2RbbhWXjPXOhBXmryc5zM9LEZiYHNwmZzaF7mXO8Oim+X0ArVowb3g0S3YXceNpyU3Of0cyNDqAmCPm0/92zsAWikAjMwaG49dsOkitElx/WjLa9maK62QiTD48cM5A4LAPzZHXxMIxsSSG+B41o6JCxyBJEltzq+gX3r7fOcSo52/nDGTnoWru+WIn/1qa1rJTeQZk/gapM9u17YL6Ap7Y9CQ/Zi/j/NQFzE6c1UIRAECloWDcIozFe0n98WHU1nY4NR4HrVrLTcNv4sVtL7KuYF2HbVeh8+jUaQIhxCTgUUmSZnu+PwAgSdITrfVvq2k6o6SOjZ7qbJNSQhgZF4hPsxoCGSV17CuUC8vsL64jxKhjfGIw8SG+bMutZlteFf3CjYxLCia2nRp9L6fXTRM0sj6znA1ZFbjcEmMSg8ktryc2yI/RCYEE+x09pCurrJ4NWRUU1ViZlBzCyPhAfHVHN4jtK6xhfVYFVrubSakhDI8JaLNjWDs4adN0g83J9rwqNmZXEhtoIDnMj+2HqimvszMg0p8gg5ZB0aa+NMXVWZzUWOwtrOHP723m+YtHnnBsfYPNySPf7+W+swcwf2SM3Oi0wXtzIWo4DD6/Tdspt1SwNHspm4s3MTl6MiPDR6Fqy+G5nYSlLcO/eA95U26jKnnayecz8JBRlcGrO17l1pG3csnAS1Ad2/rQc8xvpyCdrQwsBM6WJOk6z/crgQmSJN3WrM8NwA0A8fHxY3Jze2Q9pL7CMS82ZSy6jOPe9JSx6DJOeCzqrA7Of2Ud/cKNzBsRfaxNHJfcigaeWn6A20+L53zdJlJ2Pg2+wTDtL14iunCTV5uHw+3A6rJRa6uhuKGY9Kp0yizl9A/qz8iwka1bAo6Df1Uu8ft/QkJQlDyV2tBUbL6BWIwRWEyRJ3xsRQ1FvLf3PcwOM+ckncPgkMGE+4bTL6gfMcaY5l0VZaAb6XZl4Ij+ZUBn3fVCgZ7s3toV8pVLknR2WzqexFj09N/5ePSocYBOvy6ge8asq/d5tP2d8Fj4DZkRGDr3nhSXpc6NJLUsEdlO1L4mDYARC3t8rqXC7Ha2uDtrVEIyqFoPbZI8n5NABYS0ciiDQqKcJ2UtEKAxarxMcrZimyXjbxn7mjW1aywUOpbOdiAsAJpP3sd62lpFkqR25OdsH0KILZIkje2s7Z8sPU2+Ex2LnnYc7aUnyt+Z1wV0zzF39T47an+tjUXnHMtCQjp2g+2m+XG14tHQMdzfWRtWaC+d7Rm1GegnhEgSQuiAS4HvO3mfCgoKCgoKCu2gUy0DkiQ5hRC3AcuRQwvfkSRpb2fuU0FBQUFBQaF9dHqeAUmSlgHLOns/beCN7hbgOPR0+dpKbz+O3i7/idAdx9zV++zM/fXVc6avHpdCK/SoDIQKCgoKCgoKXU/PyqaioKCgoKCg0OUoyoCCgoKCgsIpTp9TBoQQOUKI3UKIHUKIFin0hMxLQoiDQohdQojRXSjbAI9cjZ9aIcRdR/Q5XQhR06zPw10l38kghIgTQqwUQuwTQuwVQtzZ3TK1FyGEWgixXQixpLtl6WjaMj4dfe519bXY1deXEOJsIcQBj/x/OxnZexJ94VpWaD+d7kDYTZwhSdLREpqcA/TzfCYAr3n+djqSJB0ARkJT3YYC4JtWuv4uSdLcrpCpA3EC90qStE0I4Q9sFUL8IknSvuOt2IO4Ezmk2nS8jr2Qto5PR597XXYtduX11azuykwgH9gshPi+l53vR6MvXMsK7aTPWQbawHzgA0lmAxAohIjqBjnOBDIlSeoTeWYlSSqSJGmb5/865IdqzLHX6jkIIWKBOcBb3S1LZ9BDx6czr8XOvr7GAwclScqSJMkOfIZ8PL2eHnquKHQyfVEZkICfhRBbPTnFjyQGONTsez7dc6JfCnx6lGWThBA7hRA/CiGGdKVQHYEQIhEYBWzsZlHawwvAfcBJp5Xt6RxnfDry3OvOa7Gzr6+ech/pVHrptaxwAvTFaYKpkiQVCCHCgV+EEPslSVrT3UI1x5ON8TzggVYWbwMSJEmqF0KcC3yLbEbtFQghjMBi4C5JkjquJmonIoSYC5RKkrRVCHF6N4vTqRxnfDr63OuWa7EvX19dSW+8lhVOnD5nGZAkqcDztxR5vnD8EV3aVS+hkzgH2CZJUsmRCyRJqpUkqd7z/zJAK4QI7WL5TgghhBb55vGxJElfd7c87WAKcJ4QIgfZ3DtDCPFR94rU8RxvfDr63OvGa7Errq+ecB/pNHrxtaxwgvQpZUAI4edxeEEI4QfMAvYc0e174CqPJ/NEoEaSpKIuFvUyjmLCFEJECiGXBxNCjEceo4oulO2E8Mj8NpAmSdJz3S1Pe5Ak6QFJkmIlSUpENi//JknSFd0sVofSlvHpyHOvm6/Frri++mzdld58LSucOH1tmiAC+MZzrWuATyRJ+kkIcROAJEmvI6dGPhc4CJiBa7pSQM+NcSZwY7O25vItBG4WQjgBC3Cp1DvSRE4BrgR2CyF2eNoe9Lx9KXQ/rY4PEA+dcu51y7XYVddXH6+7olzLpyBKOmIFBQUFBYVTnD41TaCgoKCgoKDQfhRlQEFBQUFB4RRHUQYUFBQUFBROcRRlQEFBQUFB4RRHUQYUFBQUFBROcRRl4Bh4KpwdtYKdEGKREOK/nbDfRUKI6Gbfc3pL4qHO5nhj0ob1xwohXjrKshwhRKgQIlAIcUtH7bM3cOQ5d4x+7wkhFh5j+SohxNgOlu2UGw/ouDFpw/r/EEKc1Up70+/s+X9yR+1ToeehKAM9k0XAcW8CCu1HkqQtkiTdcZxugcAtx+nT11hEzz3nAjn1xgO6aEwkSXpYkqRfj9PtdGDycfoo9GJ6vTLgyXS21FN4ZI8Q4hIhxBghxGohF0hZLjyV0DxvLS8KuY75Hk8GMoQQ44UQ64Vcy36dEGLACcgRJoRYLITY7PlM8bQ/KoR4x7PvLCHEHc3W+buQ66H/IYT4VAjxF4+2PRb42COnwdP9diHENiHXhx940j9cJ9KdY+L5fQKFTIUQ4ipP+wdCiJlHvO2ECCF+FnLN9rcA4dnMk0CKR6anPW1GIcRXQoj9QoiPhRCi5d57DkKIxGaypnlk921tHFo754QQD3vO4z1CiDdO5HiFELM8Y7hNCPGlkHPdN1pgHjvyfPZcQ780jocQIlfIFrFePx7QPWMihBgnhPja8/98IYRFCKETQvgIIbI87U1v+UKIsz0ybgMuaJQbuAm42yPLNM/mT/Ncm1lCsRL0fiRJ6tUf4ELgzWbfA4B1QJjn+yXI2cEAVjX2BU4D9nj+NwEaz/9nAYs9/58OLDnGvhcB//X8/wlyYRaQs7qlef5/1COPHghFTn2qBcYBOwAfwB/IAP7STM6xzfaTA9zu+f8W4K3u/t178Ji8jlyKeChyytjGbWcAfs3XB14CHvb8Pwe5yl4okNgoR7N91iDnn1cB6xvHuqd+PMcgAVM8398B/nqccWh+zgU3+/9DYJ7n//eAhcfY7yrkh1gosAbw87Tf3+y3bvV8Bv4LPOD5/+y+NB7dNSbI2R+zPP8/47kmpgDTgU+br498LzqEXLhJAF80u1YexXN/arbOl57ffzByOedu/42Vz4l/+kI64t3As0KI/wBLgCrkB8EvHsVZDTTPd/4pgCRJa4QQJiFEIPLD+H0hRD/ki1V7AnKcBQxupqybGt+EgKWSJNkAmxCiFDlV6xTgO0mSrIBVCPHDcbbfWCxkKx6NvQfTnWPyO7JSkQu8BtwghIgBqiRJajjiZeo0PL+lJElLhRBVx9juJkmS8gGEnKI1EfijjTJ1F4ckSVrr+f8j5PTDxxqH5pwhhLgP8AWCgb3A8c7R5kxEfkis9exLh/zQbqS183kqsABAklMX97XxgC4eE0lOm5wphBiEXCjqOeTzXo18rTRnIJAtSVIGgJCLdbVWerqRbyVJcgP7hBARx5JDoefT65UBSZLShRCjkXOc/wv4DdgrSdKko63Syvd/AislSVrgMYmtOgFRVMBEz8O9Cc8FbmvW5OLEfvfGbZzo+l1GN4/JGuBWZOvM/yE/XBbS8sbXXjpiDLuaI3/XOo49DgAIIXyAV5HfSg8JIR5FfmtsDwL4RZKky46y/GTP5944HtA9Y7IGuZKjA/gV+a1ejWyVOBmaj0GPn6ZRODZ9wWcgGjBLkvQR8DQwAQgTQkzyLNcKIYY0W+UST/tU5CppNchm7Mbyo4tOUJSfgdubyTXyOP3XAvM8c3dGYG6zZXXIb8a9ku4cE0mSDiGblvtJkpSF/Lb4F+Qb4pGsAS737PscIMjT3qt//2bEN/7myMe5gaOPQ/NjbnzIlHvOzROZD94ATBFCpHr25SeE6H+cddYCF3v6z6LvjQd0z5j8DtwFrJckqQwIAQbQsorkfiBRCJHi+d5cketLY6DQCr1eGQCGAZs8psJHgIeRL5T/CCF2Is/LN/eCtQohtiPPLV/raXsKeMLTfqJvGHcAY4UQu4QQ+5Adbo6KJEmbkUue7gJ+RDat13gWvwe8LrwdCHsT3T0mG4F0z/+/AzG0bkJ+DNkJai+yqToPQJKkCmTz9h5x2GGtN3IAuFUIkYb8YH2Zo4/De3jOOeQ3vjeRHxbLkeeZ24XnobMI+FQIsQt5iuB4jq+PAbOEEHuAi4BioK4PjQd0z5hsRJ6abFSIdwG7JUnyslJ4rJo3AEs9DoSlzRb/ACw4woFQoQ9xSlUtFEKsQnaC2dLdsgAIIYySJNULIXyRL9QbJEna1t1ydSU9bUz6Cp6plSWSJA3tblnaihBCD7g889yTgNckSRrZzWJ1GL1xTBROHXrLPFtf5Q0hxGBkE+D7p5oioKBwBPHAF0IIFWAHru9meRQUThlOKcvAiSKEuAa484jmtZIk3dod8igoY9LdCCG+AZKOaL5fkqTl3SGPgjImCieHogwoKCgoKCic4vQFB0IFBQUFBQWFk0BRBhQUFBQUFE5xFGVAQUFBQUHhFEdRBhQUFBQUFE5xFGVAQUFBQUHhFOf/AUDXAJWjs9IxAAAAAElFTkSuQmCC", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "%matplotlib inline\n", "import seaborn as sns\n", "sns.pairplot(iris, hue='species', height=1.5);" ] }, { "attachments": {}, "cell_type": "markdown", "metadata": { "deletable": true, "editable": true }, "source": [ "For use in Scikit-Learn, we will extract the features matrix and target array from the `DataFrame`, which we can do using some of the Pandas `DataFrame` operations discussed in [Part 3](03.00-Introduction-to-Pandas.ipynb):" ] }, { "cell_type": "code", "execution_count": 3, "metadata": { "collapsed": false, "deletable": true, "editable": true, "jupyter": { "outputs_hidden": false } }, "outputs": [ { "data": { "text/plain": [ "(150, 4)" ] }, "execution_count": 3, "metadata": {}, "output_type": "execute_result" } ], "source": [ "X_iris = iris.drop('species', axis=1)\n", "X_iris.shape" ] }, { "cell_type": "code", "execution_count": 4, "metadata": { "collapsed": false, "deletable": true, "editable": true, "jupyter": { "outputs_hidden": false } }, "outputs": [ { "data": { "text/plain": [ "(150,)" ] }, "execution_count": 4, "metadata": {}, "output_type": "execute_result" } ], "source": [ "y_iris = iris['species']\n", "y_iris.shape" ] }, { "cell_type": "markdown", "metadata": { "deletable": true, "editable": true }, "source": [ "To summarize, the expected layout of features and target values is visualized in the following figure." ] }, { "cell_type": "markdown", "metadata": { "deletable": true, "editable": true }, "source": [ "![](images/05.02-samples-features.png)\n", "[figure source in Appendix](https://github.com/jakevdp/PythonDataScienceHandbook/blob/master/notebooks/06.00-Figure-Code.ipynb#Features-and-Labels-Grid)" ] }, { "cell_type": "markdown", "metadata": { "deletable": true, "editable": true }, "source": [ "With this data properly formatted, we can move on to consider Scikit-Learn's Estimator API." ] }, { "cell_type": "markdown", "metadata": { "deletable": true, "editable": true }, "source": [ "## The Estimator API" ] }, { "cell_type": "markdown", "metadata": { "deletable": true, "editable": true }, "source": [ "The Scikit-Learn API is designed with the following guiding principles in mind, as outlined in the [Scikit-Learn API paper](http://arxiv.org/abs/1309.0238):\n", "\n", "- *Consistency*: All objects share a common interface drawn from a limited set of methods, with consistent documentation.\n", "\n", "- *Inspection*: All specified parameter values are exposed as public attributes.\n", "\n", "- *Limited object hierarchy*: Only algorithms are represented by Python classes; datasets are represented\n", " in standard formats (NumPy arrays, Pandas `DataFrame` objects, SciPy sparse matrices) and parameter\n", " names use standard Python strings.\n", "\n", "- *Composition*: Many machine learning tasks can be expressed as sequences of more fundamental algorithms,\n", " and Scikit-Learn makes use of this wherever possible.\n", "\n", "- *Sensible defaults*: When models require user-specified parameters, the library defines an appropriate default value.\n", "\n", "In practice, these principles make Scikit-Learn very easy to use, once the basic principles are understood.\n", "Every machine learning algorithm in Scikit-Learn is implemented via the Estimator API, which provides a consistent interface for a wide range of machine learning applications." ] }, { "cell_type": "markdown", "metadata": { "deletable": true, "editable": true }, "source": [ "### Basics of the API\n", "\n", "Most commonly, the steps in using the Scikit-Learn Estimator API are as follows:\n", "\n", "1. Choose a class of model by importing the appropriate estimator class from Scikit-Learn.\n", "2. Choose model hyperparameters by instantiating this class with desired values.\n", "3. Arrange data into a features matrix and target vector, as outlined earlier in this chapter.\n", "4. Fit the model to your data by calling the `fit` method of the model instance.\n", "5. Apply the model to new data:\n", " - For supervised learning, often we predict labels for unknown data using the `predict` method.\n", " - For unsupervised learning, we often transform or infer properties of the data using the `transform` or `predict` method.\n", "\n", "We will now step through several simple examples of applying supervised and unsupervised learning methods." ] }, { "cell_type": "markdown", "metadata": { "deletable": true, "editable": true }, "source": [ "### Supervised Learning Example: Simple Linear Regression\n", "\n", "As an example of this process, let's consider a simple linear regression—that is, the common case of fitting a line to $(x, y)$ data.\n", "We will use the following simple data for our regression example (see the following figure):" ] }, { "cell_type": "code", "execution_count": 5, "metadata": { "collapsed": false, "deletable": true, "editable": true, "jupyter": { "outputs_hidden": false } }, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "import matplotlib.pyplot as plt\n", "import numpy as np\n", "\n", "rng = np.random.RandomState(42)\n", "x = 10 * rng.rand(50)\n", "y = 2 * x - 1 + rng.randn(50)\n", "plt.scatter(x, y);" ] }, { "cell_type": "markdown", "metadata": { "deletable": true, "editable": true }, "source": [ "With this data in place, we can use the recipe outlined earlier. Let's walk through the process: " ] }, { "cell_type": "markdown", "metadata": { "deletable": true, "editable": true }, "source": [ "#### 1. Choose a class of model\n", "\n", "In Scikit-Learn, every class of model is represented by a Python class.\n", "So, for example, if we would like to compute a simple `LinearRegression` model, we can import the linear regression class:" ] }, { "cell_type": "code", "execution_count": 6, "metadata": { "deletable": true, "editable": true, "tags": [] }, "outputs": [], "source": [ "from sklearn.linear_model import LinearRegression" ] }, { "cell_type": "markdown", "metadata": { "deletable": true, "editable": true }, "source": [ "Note that other more general linear regression models exist as well; you can read more about them in the [`sklearn.linear_model` module documentation](http://Scikit-Learn.org/stable/modules/linear_model.html)." ] }, { "cell_type": "markdown", "metadata": { "deletable": true, "editable": true }, "source": [ "#### 2. Choose model hyperparameters\n", "\n", "An important point is that *a class of model is not the same as an instance of a model*.\n", "\n", "Once we have decided on our model class, there are still some options open to us.\n", "Depending on the model class we are working with, we might need to answer one or more questions like the following:\n", "\n", "- Would we like to fit for the offset (i.e., *y*-intercept)?\n", "- Would we like the model to be normalized?\n", "- Would we like to preprocess our features to add model flexibility?\n", "- What degree of regularization would we like to use in our model?\n", "- How many model components would we like to use?\n", "\n", "These are examples of the important choices that must be made *once the model class is selected*.\n", "These choices are often represented as *hyperparameters*, or parameters that must be set before the model is fit to data.\n", "In Scikit-Learn, hyperparameters are chosen by passing values at model instantiation.\n", "We will explore how you can quantitatively choose hyperparameters in [Hyperparameters and Model Validation](05.03-Hyperparameters-and-Model-Validation.ipynb).\n", "\n", "For our linear regression example, we can instantiate the `LinearRegression` class and specify that we would like to fit the intercept using the `fit_intercept` hyperparameter:" ] }, { "cell_type": "code", "execution_count": 7, "metadata": { "collapsed": false, "deletable": true, "editable": true, "jupyter": { "outputs_hidden": false } }, "outputs": [ { "data": { "text/plain": [ "LinearRegression()" ] }, "execution_count": 7, "metadata": {}, "output_type": "execute_result" } ], "source": [ "model = LinearRegression(fit_intercept=True)\n", "model" ] }, { "cell_type": "markdown", "metadata": { "deletable": true, "editable": true }, "source": [ "Keep in mind that when the model is instantiated, the only action is the storing of these hyperparameter values.\n", "In particular, we have not yet applied the model to any data: the Scikit-Learn API makes very clear the distinction between *choice of model* and *application of model to data*." ] }, { "cell_type": "markdown", "metadata": { "deletable": true, "editable": true }, "source": [ "#### 3. Arrange data into a features matrix and target vector\n", "\n", "Previously we examined the Scikit-Learn data representation, which requires a two-dimensional features matrix and a one-dimensional target array.\n", "Here our target variable `y` is already in the correct form (a length-`n_samples` array), but we need to massage the data `x` to make it a matrix of size `[n_samples, n_features]`.\n", "In this case, this amounts to a simple reshaping of the one-dimensional array:" ] }, { "cell_type": "code", "execution_count": 8, "metadata": { "collapsed": false, "deletable": true, "editable": true, "jupyter": { "outputs_hidden": false } }, "outputs": [ { "data": { "text/plain": [ "(50, 1)" ] }, "execution_count": 8, "metadata": {}, "output_type": "execute_result" } ], "source": [ "X = x[:, np.newaxis]\n", "X.shape" ] }, { "cell_type": "markdown", "metadata": { "deletable": true, "editable": true }, "source": [ "#### 4. Fit the model to the data\n", "\n", "Now it is time to apply our model to the data.\n", "This can be done with the `fit` method of the model:" ] }, { "cell_type": "code", "execution_count": 9, "metadata": { "collapsed": false, "deletable": true, "editable": true, "jupyter": { "outputs_hidden": false } }, "outputs": [ { "data": { "text/plain": [ "LinearRegression()" ] }, "execution_count": 9, "metadata": {}, "output_type": "execute_result" } ], "source": [ "model.fit(X, y)" ] }, { "cell_type": "markdown", "metadata": { "deletable": true, "editable": true }, "source": [ "This `fit` command causes a number of model-dependent internal computations to take place, and the results of these computations are stored in model-specific attributes that the user can explore.\n", "In Scikit-Learn, by convention all model parameters that were learned during the `fit` process have trailing underscores; for example in this linear model, we have the following:" ] }, { "cell_type": "code", "execution_count": 10, "metadata": { "collapsed": false, "deletable": true, "editable": true, "jupyter": { "outputs_hidden": false } }, "outputs": [ { "data": { "text/plain": [ "array([1.9776566])" ] }, "execution_count": 10, "metadata": {}, "output_type": "execute_result" } ], "source": [ "model.coef_" ] }, { "cell_type": "code", "execution_count": 11, "metadata": { "collapsed": false, "deletable": true, "editable": true, "jupyter": { "outputs_hidden": false } }, "outputs": [ { "data": { "text/plain": [ "-0.9033107255311146" ] }, "execution_count": 11, "metadata": {}, "output_type": "execute_result" } ], "source": [ "model.intercept_" ] }, { "attachments": {}, "cell_type": "markdown", "metadata": { "deletable": true, "editable": true }, "source": [ "These two parameters represent the slope and intercept of the simple linear fit to the data.\n", "Comparing the results to the data definition, we see that they are close to the values used to generate the data: a slope of 2 and intercept of –1.\n", "\n", "One question that frequently comes up regards the uncertainty in such internal model parameters.\n", "In general, Scikit-Learn does not provide tools to draw conclusions from internal model parameters themselves: interpreting model parameters is much more a *statistical modeling* question than a *machine learning* question.\n", "Machine learning instead focuses on what the model *predicts*.\n", "If you would like to dive into the meaning of fit parameters within the model, other tools are available, including the [`statsmodels` Python package](http://statsmodels.sourceforge.net/)." ] }, { "cell_type": "markdown", "metadata": { "deletable": true, "editable": true }, "source": [ "#### 5. Predict labels for unknown data\n", "\n", "Once the model is trained, the main task of supervised machine learning is to evaluate it based on what it says about new data that was not part of the training set.\n", "In Scikit-Learn, this can be done using the `predict` method.\n", "For the sake of this example, our \"new data\" will be a grid of *x* values, and we will ask what *y* values the model predicts:" ] }, { "cell_type": "code", "execution_count": 12, "metadata": { "deletable": true, "editable": true, "tags": [] }, "outputs": [], "source": [ "xfit = np.linspace(-1, 11)" ] }, { "cell_type": "markdown", "metadata": { "deletable": true, "editable": true }, "source": [ "As before, we need to coerce these *x* values into a `[n_samples, n_features]` features matrix, after which we can feed it to the model:" ] }, { "cell_type": "code", "execution_count": 13, "metadata": { "collapsed": false, "deletable": true, "editable": true, "jupyter": { "outputs_hidden": false } }, "outputs": [], "source": [ "Xfit = xfit[:, np.newaxis]\n", "yfit = model.predict(Xfit)" ] }, { "cell_type": "markdown", "metadata": { "deletable": true, "editable": true }, "source": [ "Finally, let's visualize the results by plotting first the raw data, and then this model fit (see the following figure):" ] }, { "cell_type": "code", "execution_count": 14, "metadata": { "collapsed": false, "deletable": true, "editable": true, "jupyter": { "outputs_hidden": false } }, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "plt.scatter(x, y)\n", "plt.plot(xfit, yfit);" ] }, { "attachments": {}, "cell_type": "markdown", "metadata": { "deletable": true, "editable": true }, "source": [ "Typically the efficacy of the model is evaluated by comparing its results to some known baseline, as we will see in the next example." ] }, { "cell_type": "markdown", "metadata": { "deletable": true, "editable": true }, "source": [ "### Supervised Learning Example: Iris Classification\n", "\n", "Let's take a look at another example of this process, using the Iris dataset we discussed earlier.\n", "Our question will be this: given a model trained on a portion of the Iris data, how well can we predict the remaining labels?\n", "\n", "For this task, we will use a simple generative model known as *Gaussian naive Bayes*, which proceeds by assuming each class is drawn from an axis-aligned Gaussian distribution (see [In Depth: Naive Bayes Classification](05.05-Naive-Bayes.ipynb) for more details).\n", "Because it is so fast and has no hyperparameters to choose, Gaussian naive Bayes is often a good model to use as a baseline classification, before exploring whether improvements can be found through more sophisticated models.\n", "\n", "We would like to evaluate the model on data it has not seen before, so we will split the data into a *training set* and a *testing set*.\n", "This could be done by hand, but it is more convenient to use the `train_test_split` utility function:" ] }, { "cell_type": "code", "execution_count": 15, "metadata": { "deletable": true, "editable": true, "tags": [] }, "outputs": [], "source": [ "from sklearn.model_selection import train_test_split\n", "Xtrain, Xtest, ytrain, ytest = train_test_split(X_iris, y_iris,\n", " random_state=1)" ] }, { "cell_type": "markdown", "metadata": { "deletable": true, "editable": true }, "source": [ "With the data arranged, we can follow our recipe to predict the labels:" ] }, { "cell_type": "code", "execution_count": 16, "metadata": { "collapsed": false, "deletable": true, "editable": true, "jupyter": { "outputs_hidden": false } }, "outputs": [], "source": [ "from sklearn.naive_bayes import GaussianNB # 1. choose model class\n", "model = GaussianNB() # 2. instantiate model\n", "model.fit(Xtrain, ytrain) # 3. fit model to data\n", "y_model = model.predict(Xtest) # 4. predict on new data" ] }, { "cell_type": "markdown", "metadata": { "deletable": true, "editable": true }, "source": [ "Finally, we can use the ``accuracy_score`` utility to see the fraction of predicted labels that match their true values:" ] }, { "cell_type": "code", "execution_count": 17, "metadata": { "collapsed": false, "deletable": true, "editable": true, "jupyter": { "outputs_hidden": false } }, "outputs": [ { "data": { "text/plain": [ "0.9736842105263158" ] }, "execution_count": 17, "metadata": {}, "output_type": "execute_result" } ], "source": [ "from sklearn.metrics import accuracy_score\n", "accuracy_score(ytest, y_model)" ] }, { "cell_type": "markdown", "metadata": { "deletable": true, "editable": true }, "source": [ "With an accuracy topping 97%, we see that even this very naive classification algorithm is effective for this particular dataset!" ] }, { "cell_type": "markdown", "metadata": { "deletable": true, "editable": true }, "source": [ "### Unsupervised Learning Example: Iris Dimensionality\n", "\n", "As an example of an unsupervised learning problem, let's take a look at reducing the dimensionality of the Iris data so as to more easily visualize it.\n", "Recall that the Iris data is four-dimensional: there are four features recorded for each sample.\n", "\n", "The task of dimensionality reduction centers around determining whether there is a suitable lower-dimensional representation that retains the essential features of the data.\n", "Often dimensionality reduction is used as an aid to visualizing data: after all, it is much easier to plot data in two dimensions than in four dimensions or more!\n", "\n", "Here we will use *principal component analysis* (PCA; see [In Depth: Principal Component Analysis](05.09-Principal-Component-Analysis.ipynb)), which is a fast linear dimensionality reduction technique.\n", "We will ask the model to return two components—that is, a two-dimensional representation of the data.\n", "\n", "Following the sequence of steps outlined earlier, we have:" ] }, { "cell_type": "code", "execution_count": 18, "metadata": { "deletable": true, "editable": true, "tags": [] }, "outputs": [], "source": [ "from sklearn.decomposition import PCA # 1. Choose the model class\n", "model = PCA(n_components=2) # 2. Instantiate the model\n", "model.fit(X_iris) # 3. Fit to data\n", "X_2D = model.transform(X_iris) # 4. Transform the data" ] }, { "cell_type": "markdown", "metadata": { "deletable": true, "editable": true }, "source": [ "Now let's plot the results. A quick way to do this is to insert the results into the original Iris `DataFrame`, and use Seaborn's `lmplot` to show the results (see the following figure):" ] }, { "cell_type": "code", "execution_count": 19, "metadata": { "collapsed": false, "deletable": true, "editable": true, "jupyter": { "outputs_hidden": false } }, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "iris['PCA1'] = X_2D[:, 0]\n", "iris['PCA2'] = X_2D[:, 1]\n", "sns.lmplot(x=\"PCA1\", y=\"PCA2\", hue='species', data=iris, fit_reg=False);" ] }, { "cell_type": "markdown", "metadata": { "deletable": true, "editable": true }, "source": [ "We see that in the two-dimensional representation, the species are fairly well separated, even though the PCA algorithm had no knowledge of the species labels!\n", "This suggests to us that a relatively straightforward classification will probably be effective on the dataset, as we saw before." ] }, { "cell_type": "markdown", "metadata": { "deletable": true, "editable": true }, "source": [ "### Unsupervised Learning Example: Iris Clustering\n", "\n", "Let's next look at applying clustering to the Iris data.\n", "A clustering algorithm attempts to find distinct groups of data without reference to any labels.\n", "Here we will use a powerful clustering method called a *Gaussian mixture model* (GMM), discussed in more detail in [In Depth: Gaussian Mixture Models](05.12-Gaussian-Mixtures.ipynb).\n", "A GMM attempts to model the data as a collection of Gaussian blobs.\n", "\n", "We can fit the Gaussian mixture model as follows:" ] }, { "cell_type": "code", "execution_count": 20, "metadata": { "collapsed": false, "deletable": true, "editable": true, "jupyter": { "outputs_hidden": false } }, "outputs": [], "source": [ "from sklearn.mixture import GaussianMixture # 1. Choose the model class\n", "model = GaussianMixture(n_components=3,\n", " covariance_type='full') # 2. Instantiate the model\n", "model.fit(X_iris) # 3. Fit to data\n", "y_gmm = model.predict(X_iris) # 4. Determine labels" ] }, { "cell_type": "markdown", "metadata": { "deletable": true, "editable": true }, "source": [ "As before, we will add the cluster label to the Iris ``DataFrame`` and use Seaborn to plot the results (see the following figure):" ] }, { "cell_type": "code", "execution_count": 21, "metadata": { "collapsed": false, "deletable": true, "editable": true, "jupyter": { "outputs_hidden": false } }, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "iris['cluster'] = y_gmm\n", "sns.lmplot(x=\"PCA1\", y=\"PCA2\", data=iris, hue='species',\n", " col='cluster', fit_reg=False);" ] }, { "cell_type": "markdown", "metadata": { "deletable": true, "editable": true }, "source": [ "By splitting the data by cluster number, we see exactly how well the GMM algorithm has recovered the underlying labels: the *setosa* species is separated perfectly within cluster 0, while there remains a small amount of mixing between *versicolor* and *virginica*.\n", "This means that even without an expert to tell us the species labels of the individual flowers, the measurements of these flowers are distinct enough that we could *automatically* identify the presence of these different groups of species with a simple clustering algorithm!\n", "This sort of algorithm might further give experts in the field clues as to the relationships between the samples they are observing." ] }, { "cell_type": "markdown", "metadata": { "deletable": true, "editable": true }, "source": [ "## Application: Exploring Handwritten Digits" ] }, { "cell_type": "markdown", "metadata": { "deletable": true, "editable": true }, "source": [ "To demonstrate these principles on a more interesting problem, let's consider one piece of the optical character recognition problem: the identification of handwritten digits.\n", "In the wild, this problem involves both locating and identifying characters in an image. Here we'll take a shortcut and use Scikit-Learn's set of preformatted digits, which is built into the library." ] }, { "cell_type": "markdown", "metadata": { "deletable": true, "editable": true }, "source": [ "### Loading and Visualizing the Digits Data\n", "\n", "We can use Scikit-Learn's data access interface to take a look at this data:" ] }, { "cell_type": "code", "execution_count": 22, "metadata": { "collapsed": false, "deletable": true, "editable": true, "jupyter": { "outputs_hidden": false } }, "outputs": [ { "data": { "text/plain": [ "(1797, 8, 8)" ] }, "execution_count": 22, "metadata": {}, "output_type": "execute_result" } ], "source": [ "from sklearn.datasets import load_digits\n", "digits = load_digits()\n", "digits.images.shape" ] }, { "cell_type": "markdown", "metadata": { "deletable": true, "editable": true }, "source": [ "The images data is a three-dimensional array: 1,797 samples each consisting of an 8 × 8 grid of pixels.\n", "Let's visualize the first hundred of these (see the following figure):" ] }, { "cell_type": "code", "execution_count": 23, "metadata": { "collapsed": false, "deletable": true, "editable": true, "jupyter": { "outputs_hidden": false } }, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "import matplotlib.pyplot as plt\n", "\n", "fig, axes = plt.subplots(10, 10, figsize=(8, 8),\n", " subplot_kw={'xticks':[], 'yticks':[]},\n", " gridspec_kw=dict(hspace=0.1, wspace=0.1))\n", "\n", "for i, ax in enumerate(axes.flat):\n", " ax.imshow(digits.images[i], cmap='binary', interpolation='nearest')\n", " ax.text(0.05, 0.05, str(digits.target[i]),\n", " transform=ax.transAxes, color='green')" ] }, { "cell_type": "markdown", "metadata": { "deletable": true, "editable": true }, "source": [ "In order to work with this data within Scikit-Learn, we need a two-dimensional, `[n_samples, n_features]` representation.\n", "We can accomplish this by treating each pixel in the image as a feature: that is, by flattening out the pixel arrays so that we have a length-64 array of pixel values representing each digit.\n", "Additionally, we need the target array, which gives the previously determined label for each digit.\n", "These two quantities are built into the digits dataset under the `data` and `target` attributes, respectively:" ] }, { "cell_type": "code", "execution_count": 24, "metadata": { "collapsed": false, "deletable": true, "editable": true, "jupyter": { "outputs_hidden": false } }, "outputs": [ { "data": { "text/plain": [ "(1797, 64)" ] }, "execution_count": 24, "metadata": {}, "output_type": "execute_result" } ], "source": [ "X = digits.data\n", "X.shape" ] }, { "cell_type": "code", "execution_count": 25, "metadata": { "collapsed": false, "deletable": true, "editable": true, "jupyter": { "outputs_hidden": false } }, "outputs": [ { "data": { "text/plain": [ "(1797,)" ] }, "execution_count": 25, "metadata": {}, "output_type": "execute_result" } ], "source": [ "y = digits.target\n", "y.shape" ] }, { "cell_type": "markdown", "metadata": { "deletable": true, "editable": true }, "source": [ "We see here that there are 1,797 samples and 64 features." ] }, { "cell_type": "markdown", "metadata": { "deletable": true, "editable": true }, "source": [ "### Unsupervised Learning Example: Dimensionality Reduction\n", "\n", "We'd like to visualize our points within the 64-dimensional parameter space, but it's difficult to effectively visualize points in such a high-dimensional space.\n", "Instead, we'll reduce the number of dimensions, using an unsupervised method.\n", "Here, we'll make use of a manifold learning algorithm called Isomap (see [In-Depth: Manifold Learning](05.10-Manifold-Learning.ipynb)) and transform the data to two dimensions:" ] }, { "cell_type": "code", "execution_count": 26, "metadata": { "collapsed": false, "deletable": true, "editable": true, "jupyter": { "outputs_hidden": false } }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "(1797, 2)\n" ] } ], "source": [ "from sklearn.manifold import Isomap\n", "iso = Isomap(n_components=2)\n", "iso.fit(digits.data)\n", "data_projected = iso.transform(digits.data)\n", "print(data_projected.shape)" ] }, { "cell_type": "markdown", "metadata": { "deletable": true, "editable": true }, "source": [ "We see that the projected data is now two-dimensional.\n", "Let's plot this data to see if we can learn anything from its structure (see the following figure):" ] }, { "cell_type": "code", "execution_count": 27, "metadata": { "collapsed": false, "deletable": true, "editable": true, "jupyter": { "outputs_hidden": false } }, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "plt.scatter(data_projected[:, 0], data_projected[:, 1], c=digits.target,\n", " edgecolor='none', alpha=0.5,\n", " cmap=plt.cm.get_cmap('viridis', 10))\n", "plt.colorbar(label='digit label', ticks=range(10))\n", "plt.clim(-0.5, 9.5);" ] }, { "cell_type": "markdown", "metadata": { "deletable": true, "editable": true }, "source": [ "This plot gives us some good intuition into how well various numbers are separated in the larger 64-dimensional space. For example, zeros and ones have very little overlap in the parameter space.\n", "Intuitively, this makes sense: a zero is empty in the middle of the image, while a one will generally have ink in the middle.\n", "On the other hand, there seems to be a more or less continuous spectrum between ones and fours: we can understand this by realizing that some people draw ones with \"hats\" on them, which causes them to look similar to fours.\n", "\n", "Overall, however, despite some mixing at the edges, the different groups appear to be fairly well localized in the parameter space: this suggests that even a very straightforward supervised classification algorithm should perform suitably on the full high-dimensional dataset.\n", "Let's give it a try." ] }, { "cell_type": "markdown", "metadata": { "deletable": true, "editable": true }, "source": [ "### Classification on Digits\n", "\n", "Let's apply a classification algorithm to the digits data.\n", "As we did with the Iris data previously, we will split the data into training and testing sets and fit a Gaussian naive Bayes model:" ] }, { "cell_type": "code", "execution_count": 28, "metadata": { "deletable": true, "editable": true, "tags": [] }, "outputs": [], "source": [ "Xtrain, Xtest, ytrain, ytest = train_test_split(X, y, random_state=0)" ] }, { "cell_type": "code", "execution_count": 29, "metadata": { "deletable": true, "editable": true, "tags": [] }, "outputs": [], "source": [ "from sklearn.naive_bayes import GaussianNB\n", "model = GaussianNB()\n", "model.fit(Xtrain, ytrain)\n", "y_model = model.predict(Xtest)" ] }, { "cell_type": "markdown", "metadata": { "deletable": true, "editable": true }, "source": [ "Now that we have the model's predictions, we can gauge its accuracy by comparing the true values of the test set to the predictions:" ] }, { "cell_type": "code", "execution_count": 30, "metadata": { "collapsed": false, "deletable": true, "editable": true, "jupyter": { "outputs_hidden": false } }, "outputs": [ { "data": { "text/plain": [ "0.8333333333333334" ] }, "execution_count": 30, "metadata": {}, "output_type": "execute_result" } ], "source": [ "from sklearn.metrics import accuracy_score\n", "accuracy_score(ytest, y_model)" ] }, { "cell_type": "markdown", "metadata": { "deletable": true, "editable": true }, "source": [ "With even this very simple model, we find about 83% accuracy for classification of the digits!\n", "However, this single number doesn't tell us where we've gone wrong. One nice way to do this is to use the *confusion matrix*, which we can compute with Scikit-Learn and plot with Seaborn (see the following figure):" ] }, { "cell_type": "code", "execution_count": 31, "metadata": { "collapsed": false, "deletable": true, "editable": true, "jupyter": { "outputs_hidden": false } }, "outputs": [ { "data": { "image/png": "iVBORw0KGgoAAAANSUhEUgAAAQYAAAEGCAYAAACHNTs8AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjQuMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/MnkTPAAAACXBIWXMAAAsTAAALEwEAmpwYAAAqlElEQVR4nO2dd3gU1frHP28aEBJ6CCC9BhAEBATpVUBAQKQIVgRFUUCuCKh4bfzgKtxruSAW0KuoIMoVwV7oRYr0IkUEKaG3hJJyfn/sBnPZBJLdOWQH3s/z5GF2duc7Xw7LmzMz53yPGGNQFEVJT0hOG1AUJfjQwqAoig9aGBRF8UELg6IoPmhhUBTFh7CcNpAZcSO+tfK4ZM2Lt9iQVRRXkjsMyWi/9hgURfFBC4OiKD5oYVAUxQctDIqi+KCFQVEUH7QwKIrig6sKQ0RYCDMeacB/B9/Ml0Mb8WjrCgB8+GB9Zj3WkFmPNWTBqGa8cVetgM6zeOECOt96Cx3bteHdt99ywLnq2tRUXec1g3YcQ0acT07l3rdXkHg+hbAQYdpD9Vmw9TB9J/9y4TOv9a3Fj5sO+n2OlJQUxrz0PJPfnkpsbCx39uxO8xYtqVCxYkDeVdddXt2m67Smq3oMAInnUwAICxXCQkNIPwoqb65QbqpQiB82xvutv2H9OkqVKkPJUqUIj4igXYdbmffzjwG6Vl23eXWbrtOa1gqDiMSJyJMi8pr350kRqRqobojArMcasvjpFizZdoR1e05ceK919ViWbT9CwrkUv/UPxsdTrHixC6+LxsYSH+9/oVFdu5qqa0fTSmEQkSeBTwABfvH+CPCxiIy4xHEDRGSliKw8vuarDD+TaqDra0tp/n/zqVkqP5Vioy68d+sNxZi79oCTfxVFuSaxdY+hH1DdGJOUfqeITAA2AmMzOsgY8xbwFlx+rsSps8ks33mUJpWLsC3+NAUiw6lZMj+DPlgTkPGisbEc2P9XcTkYH09sbGxAmqprT1N17WjaupRIBUpksL+49z2/KJg3nOjcnlqWKyyEmysWZuehBABuqRHLvC2HOJ/stzwA1a+vwe7du/jzzz0knT/PN1/NpVmLlgFpqq77vLpN12lNWz2GIcCPIrIN2OPdVxqoCAzyVzQmOhdje9QgVAQR+GZ9PPO2HALg1huK89a83wO0DWFhYYx8ajQDBzxAamoKXbreTsWKlVTXAV03eXWbrtOaYisMVkRCgPrAdd5de4EVxpgs3RnUadeKYp/Mpl1bG8dgjEkFltnSVxTFHq4bx6Aoin20MCiK4oMWBkVRfNDCoCiKD1oYFEXxwdrjykA5m4wVY1Ue/9KGLFsndLKiq7iL44lJl/+QHxSIDLeiqynRiqJkGS0MiqL4oIVBURQftDAoiuKDFgZFUXzQwqAoig+uLwxOJOPmCgvhi2GN+frJpnw/sjlD21cG4OZKhZn7RFO+G9GM8X1qERqS4ZOdK+rV7bpu8mpTNyUlhX59ujNi6MOOaTrp1dWFIS0Zd+Kb7zBr9ly++WoOO7Zvz7bOueRUer++lPbjFtB+3HyaVS3KjeUKMr5vbQa9t4q2Y+ez99gZutcvmeNe3azrJq82dQFmfvIhZcqVd0QLnPfq6sLgZDLuX+nTIYSHhpCSakhKSeV3b0LUwi2HaH9D8aDw6lZdN3m1qXsw/gDLFi2g4223B6yVhmtSoq8ETibjhgh8Nbwpq8e0ZeHWQ6z54zihIUKNUvkB6FCrOMUL5gkKr27VdZNXm7pvTBjHQ489jgR4aZoeV6REXwoRue8S711IiXbyei4rpBro8I8FNBj9PbXKFKBy8WgefW81o7tV54thjUk4l0xKanAOH1fcw5KF8yhQsBBVqlbPaSuXJCdWonoOmJrRG+lTorMyV8JG2u7JM8ks2XaY5lVjeOunndzx6hIAmsTFUK5o1GWOvrJe3abrJq+2dDes/ZUlC+exfMlCzp87R0JCAi8+8yRPvzAuqLzaWldiXSY/64HA/8W8OJWMWygqgnx5vOnT4SE0qRLD9vjTFI6KADxrZg5sVYFpi3bluFc367rJqy3dAYOGMnPuj0yf/R2jx7xMnXr1Ay4KNrza6jHEArcAxy7aL8ASp07iVDJu0Xy5mNC3NiEihAjMWbOPnzYeZNRtVWlVPRYR4cPFu1iy7UiOe3Wzrpu82tS1gStSokXkXWCqMWZRBu99ZIy583IaOu1acSNXy7RrKz0GY0y/S7x32aKgKErO4urHlYqi2EELg6IoPmhhUBTFBy0MiqL4oIVBURQfrrmU6L1Hz9iQ5fsdgY+hv5hetUo5rgmQOzzUiq4C8387ZEW3WeUYK7qaEq0oSpbRwqAoig9aGBRF8UELg6IoPmhhUBTFBy0MiqL4kBNBLY6yeOECxo19idSUVLrefgf9+g8ISO/P3bsY++zwC6/379vLXf0G0qVH32xrnTpykG/efpnEk8cBqNG8A3XaduXs6ZPMnTSGk4fjyVckllsfforceaP98ht/YD9/f3okR48eRhC63N6DXn3u8kvrYpxuW1uabtCd9voYNq5cQnT+gox87QMAvvrkXZZ+/yVR+QoA0LHvg1S/sWGOewWXF4a0ZNzJb08lNjaWO3t2p3mLllSoWNFvzZKly/LG1BkX9O/u1paGTf0LvJDQUJr2GkBs2UqcP5PItL8Pokz1Omxc9D2lqtamfsee/DJnOivmTqdJjwf8OkdoaBiDhw0nrmo1EhISuKd3d+o3aEj5Cv63AdhpWxuabtG9qWUHmna4nQ9fffF/9jfv1INWXQKfcOx0G7j6UsJWim8aa1ctp1iJksQWK+HX8VEFChNb1hOWEZEnkkIlSnH62GF2/rqUao1bA1CtcWt2rF7qt8ciMTHEVa0GQN68eSlbvjyHDh70Wy8NTYl2Vrdi9VpERucL2FNmuCYlWkTiRKSViERdtL+dU+ewleKbxvwfv6V56/aOaJ04dIBDf+ygWIU4Ek8cI6pAYQDy5i9E4omLg678Y9/evfy2ZTPVa9QMWEtTou1/vwAWfvU5Y4fcw7TXx5B4+qTfOq5IiRaRx4AvgEeBDSJyW7q3x1ziuBxLib6YpKQkli+eT+MWbQLWOn/2DHPeeIFmdz5Erjx5/+c9EQEJPEY8MTGBEX8bzNAnRhIV5X9orXLlaNyuK6MnTWf4hKnkL1iYWVPfyGlLF7B1j6E/cKMx5rSIlAVmikhZY8yrkPHYbAiOlOg0Vi5bRIXKcRQsVDggnZTkZOa88QJxDVtSqW5jACLzF+T08SNEFSjM6eNHiPTefPKX5KQkRgwbQrsOHWnRKvBCBpoSbVM3jXwFCl3Ybti2M2+9OPwSn740rkiJBkKMMacBjDG7gOZAexGZwCUKQ3axlQ4MMP+Hb2jWKrCrHmMM30+ZQKHipbix3V+rDpWv1YBNi34AYNOiHyhf2/870cYYXnzuGcqWK8+dd90bkN/0aEq03e8XwImjhy9sr1u2gOJl/F+yzi0p0fEiUssYswbA23PoCEwBajh1ElspvmfPnOHXlct49ImnA9LZt20jm5f8SJGS5fjwmYEANOp+H/U69mTuv19i48JviC5clI4PP+X3OdauWc3Xc2ZTsVJl+vboCsDAR4fQqEmzgLxrSrSzuu+Nf5btG9dw+uRxnnmgKx169WPbhl/Z+/s2RIRCRYvR86EngsIr2EuJLgkkG2MOZPBeI2PM4stp6LRrnXbtRq6Wade2UqL/vMR7ly0KiqLkLK4ex6Aoih20MCiK4oMWBkVRfNDCoCiKD1oYFEXx4ZpLiXYTugCvPZJT7Hy9bC1qWyQ6woqupkQripJltDAoiuKDFgZFUXzQwqAoig9aGBRF8cH1hWHxwgV0vvUWOrZrg1PhLjY0ndTNFRbCF8Ma8/WTTfl+ZHOGtq8MwM2VCjP3iaZ8N6IZ4/vUIjQksBnu12Lbpue50aNo3exmenR1/inOnV1u4YE+XRlwV3cG3tvTEU0n28DVhSEtAHPim+8wa/ZcvvlqDju2bw86Tad1zyWn0vv1pbQft4D24+bTrGpRbixXkPF9azPovVW0HTufvcfO0L1+yaDwa1PTpm6nzl15fdLbAetkxvh/T+GtD2Yy6b3pAWs53QauLgzXcmBp4vkUAMJCQwgPDSEl1ZCUksrvhxIAWLjlEO1vKB40fm1p2tStU7ce+fPnD1jnSuCmMNj6IlLPu11NRB4XkQ5OnuNaDiwNEfhqeFNWj2nLwq2HWPPHcUJDhBqlPF/kDrWKU7xgnqDxa0vTpq5NRIThjz3IQ/f0YM5/Pw1Yz+k2sJLHICLPAu2BMBH5HrgJ+BkYISK1jTEvZXLcAGAAwBsTJzu2aMjVSKqBDv9YQL48Ybz1QD0qF4/m0fdWM7pbdSLCQli45RApqdf84NGg5V+T3yemaCzHjh5h+GMDKF2mHDVr181pWxewFe3WHagF5AIOACWNMSdF5BVgOZBhYQiGMFi3BYuePJPMkm2HaV41hrd+2skdry4BoElcDOWK+p8WrW1rl5iiHn8FCxWmcbNWbNm0IaDC4JYw2GRjTIoxJhHYYYw5CWCMOQOkOnWSazWwtFBUBPnyeGp6rvAQmlSJYXv8aQpHecbTR4SFMLBVBaYt2hUUfm1q2tS1xZkziSQmJFzYXvnLEsqWD2zVLLeEwZ4XkUhvYbgxbaeI5MfBwnCtBpYWzZeLCX1rEyJCiMCcNfv4aeNBRt1WlVbVYxERPly8iyXbjgSFX5uaNnVHDX+clStXcPz4Mdq3bsaDDz9Kl27dA9Y9dvQIzz45BPA8TWjVtgP1GzYOSNMtYbC5jDHnMthfBChujFl/OQ2dXamzK22isys9XOkwWJ+i4N1/GDic0XuKogQPrh7HoCiKHbQwKIrigxYGRVF8yFJhEJHGInKfdztGRMrZtaUoSk5y2cLgHcX4JDDSuysc+NCmKUVRcpbLPq4UkTVAbWC1Maa2d986Y0xNm8ZsPa609ZgqLNSxRbyt0/Xt5VZ0Z/W/yYquDQ6cOGtFt1j+3FZ0bRFIGOx546keBkBE8jppTFGU4CMrhWGGiEwGCohIf+AHwN4kdUVRcpzLDnAyxrwiIm2Ak0AVYLQx5nvrzhRFyTGyNPLRWwi0GCjKNcJlC4OInIILNwIj8DyVSDDG5LNpTFGUnCMrlxLRadsiIsBtQAObphRFyVmyNfLRePgvcIsdO9nHTenAwZyQHB4q/Ktbdf59x/W82bMGfetdB0Cn62N5984b+HrgTeTLHficu2Bug4w4feokL4waRr9et/FA7y5sWr/WEd1gT+DOyqVEt3QvQ4C6gJ2HwNkkLRl38ttTiY2N5c6e3WneoiUVKgYWetGpc1d69OrDs0+NcMipPa9O6SalGEbM3szZ5FRCQ4RXulRj5e4TbDpwiuV/HOMfnasF5NNJr1dKF2DSv/5B3QaNeGbMeJKSkjh39kzAmjb8Oq2ZlR5Dp3Q/twCn8FxO5DhuSgd2Q0Ly2WRPhk5YiBAWIhhj2HE4kYOnzgfs02mvV0I34fQp1q9ZRbtOXQEIDw8nKjrwW2tuSODOyj2G+/xWT4eI/McYc7cTWmlklIy7ft06J0/hGLa8OqkbIvBa9+spkT83czbEs/VgQsD+0uOGNkjPgX17yV+gIONfGs3ObVupFFeNgUOGkztPZEC6Nvw6rZlpYRCR1yHzYcnGmMcucezsi3cBLUSkgPfYzpkcpynROUiqgUGfbiBvRCjPtKtMmUJ5+ONo4F1nt5KSksL237bwyOMjiKtek0n/HMf0D6Zwz4BBOW3NOpfqMawMQLcksAl4B09xETz3JsZf6qBgSIm2hZsSkhPOp7Bu70nqlsrvaGFwUxsAFCkaS0xMLHHVPdOCGrdow4wPpgSs64YE7kzvMRhj3r/Uz2V06wKrgKeAE8aYecAZY8x8Y8x8v91ehJvSgYM9ITl/7jDyRoQCEBEq1C6Vjz3Hnb3HHOxtcDGFChehSGwse/7YBcCalcspXa58wLpuSODOylOJGDzTrqsBF6aOGWMyPasxJhX4p4h86v0zPivnyi5uSgcO9oTkgpHh/K1lBUJCBBFYuP0ov/xxnM41YrmjVgkKRoYzsUcNVuw+zqvzfs9Rr1dKF+CRoSMY99xIkpOSKFaiJMOeej5gTTckcGdl2vV3wHTgb8BDwD3AIWPMk1k+icitQCNjzKisHqPTru2h06512nUagUy7LmyMeRdI8l4K3A9kq49ijJmbnaKgKErOkpXufVpQ/n7vb/59QCF7lhRFyWmyUhhe9K4gNQx4HcgHDLXqSlGUHCUrhWG5MeYEcAJoYdmPoihBQFbuMSwWke9EpJ+IFLTuSFGUHOeyhcEYUxl4GqgOrBKROSLS17ozRVFyjGwtautdlHYC0McYE2rNFfq4EtzlFSD27g+s6Mb/5y7HNU+fTXZcEyDKganpVxK/H1eKSD4RuUdEvgaWAPuB+g77UxQliMhKeVsL/Bd43hiz1K4dRVGCgawUhvImO9cbiqK4nqzcfNSioCjXGLrataIoPmhhUBTFh6w8lagsIj+KyAbv65oi8rR9a1njWk+JtuUVnPUbIsLCMbcy/W+ewbNv9G/Iov+7lcVjO/KfwU3Jmyuwx3w22vbcuXM8cHdP7unVlT53dOadN99wRBeCPyU6Kz2Gt4GReCdTGWPWAb0COqtDpCXjTnzzHWbNnss3X81hx/btAet26tyV1yc5uzynm7yC834Hto9j694TF16P/HAljUfOpdGIOew5ksCAtlWCxmsaERERvPbmFN7/ZBbvf/QZy5csYoMD8fE2/DqtmZXCEGmM+eWifXZGh2QTTYm24xWc9VuiUCS31LqO//z81xf11JmkC9t5IkIDGs1mq21FhMhIz+LuycnJJCcnIxmPB8oWbkiJzkphOCwiFfAGw4pIdzyDnLKMiDQWkcdFpK0fHjMlo2Tc+Ph4J0/hGG7yCs76HXtXXUZ/vJrUix5w/fvBhmyb1J1KxfMz+dstQeH1YlJSUrindzc6tmlCvQYNqV6jZsCaNvw6rZmVwvAIMBmIE5G9wBBg4KUOEJFf0m33B94AooFnRSTTVVxEZICIrBSRlU6uJqTkHLfUvo5DJ8+y5vejPu89MnkpVR7+jN/2naBbw7JX3lwWCA0N5f2PP2fW1z+xacN6dm7fltOWrghZWVdiJ9BaRPICIcaYU1nQDU+3PQBoY4w5JCKvAMuAsZmcS1OigwSn/DaoXJT2dUrSptZ15A4PJTpPOG893IgBExcDkGoMny3dxeCO1Zk2f0eOer0U0dH5qFO3PsuWLKJ8gPmMrk6JTkNERovIaDxBLUPTvb6krogUFJHCeCZqHQIwxiTg4P0JTYm2h1N+n5v+K9Ue/Zyag2dx/+sLWbDxAAMmLqZ87IW1kulQpyTb9p24hMqV8Xoxx44d5dSpkwCcO3uWFcuXUqZsuYB1r4qUaCD9ckS5gY7A5ssckx9PfLwARkSKG2P2i0iUd58jaEq0Ha82/QKIwKSHbiY6Tzgiwobdx3h8iv8Btba8Hjl8iBefHUVqSiqpJpWWrW+hUdPmAeteFSnRPgeI5AK+NcY0z/bJRCKBWGPMZfPHddq1u7yCTruGq2fatT9/i0g8K01lG2NMIuDfogSKolwxsrLgzHr+WsMyFIgBAl91Q1GUoCUrPYaO6baTgXhjTFAMcFIUxQ6XLAwiEornfkLcFfKjKEoQcMnHlcaYFGCriJS+Qn4URQkCsnIpURDY6B3NeOHRpTGmszVXiqLkKFlZ1LZZRvudXM4+I2w9rrSFjUeLbloo1yYF6w1yXPPYCuemULuZQB5Xdrh4ZWsRGQdYLQyKouQcWZlE1SaDfe2dNqIoSvCQaY9BRAYCDwPlRWRdureigcW2jSmKknNc6lLiI+Br4P+A9FOlTxljfOfQKopy1ZBpYUi3wnXvK2dHUZRgQFOiFUXxwfWFIdjTdtNwS5qzbV0nNbfMfY4VM0ax7JMRLJo2HIBurWuzauZTJKx6jTrVAh+Xd622rasLgxvSdtNwS5qzTV0bmu0GvEqDXmNp3OcfAGzcsY9ew95m0Wr/0qBs+7WlmxMp0UGLG9J203BDmrNtXVte07P193i2/XHQEa1ruW2tFAYRuUlE8nm384jIcyLypYiMExHH/ne4IW3XNrb8uqFtjTF8OXEQi6cN5/5ujQLylhHXctva6jFMARK926/iiXob5903NbODNCVayQ6t7vsnN985ji6DJvJgzyY0qlMhpy1dNdjKoQpJl9lQ1xhTx7u9SETWZHZQMKREX6tpzldC12nNfYc8AbKHjp1m9k/rqFe9LIsduLeQxrXctrZ6DBtE5D7v9loRqQuedTDxLnXnBG5I27WNLb/B3raRuSOIisx1Ybt1wzg27tgXkL+LuVbbFuz1GB4AXvUufnsYWCoie4A93vccwQ1pu2m4Lc052Nu2aOFopk/o79ENDWX61yv5fslmOreoyYQn76BIwSg+f+0h1m3dS+dH/p3jfm3r5nhKdLbEPTcgy+EpQH8aY7J8N0SnXeu06zR02rU9nEyJzjLGmJNA4MsDK4pyRXH1OAZFUeyghUFRFB+0MCiK4oMWBkVRfNDCoCiKD1YfVwaC2x5Xugm3LZZrg4IdXraie+yrJ6zo2iKzx5XaY1AUxQctDIqi+KCFQVEUH7QwKIrigxYGRVF8cH1hCPZQTbfq2gqvdUMbhIQISyfezWfPdwPgh/G9WTbpHpZNuoedHw9kxt+7BJVfG5quLgxuCNV0q66N8Fq3tMGgrjeydfeRC69bD/uYBgPfp8HA91m+aR//XbQtqPza0HR1YXBDqKZbdW2E17qhDa4rEkW7+uWZ+s16n/eiIyNoVqs0Xy4JrDC44XtrKwz2MREpZUM7PW4I1XSrrg3c0AYvD2zJU+/MJzXVdxBYp5srMW/NH5xKPO+3V3DH99ZWj+EFYLmILBSRh0UkJisHaRiskpO0v6k8B48n8uu2jP9D9WgRx4yft1xhVzmDraCWncCNQGugJ/CciKwCPgY+N8acyuigqzkM1m26Ngj2NmhY/To6NqhIu3rlyRURRr7ICKY8eSv3j5tL4Xx5qFulOD3//t+g8WtT01aPwRhjUo0x3xlj+gElgIlAOzxFwxHcEKrpVl0bBHsbjJ6ykIp93iTu7re4e8yXzFuzm/vHzQWga5PKfL18B+eSUoLGr01NWz2G/5mYYYxJAmYDs0Uk0qmTuCFU0626NsJr3dYG6bmjeRyvTF/uiJYbvrdWZleKSGVjzG+BaOjsSnvo7EqdXZnGFZ1dGWhRUBQlZ3H1OAZFUeyghUFRFB+0MCiK4oMWBkVRfNDCoCiKD9dcGOyBE2dtyFIsf27HNd32WNFNfs86MFApI258+jsruhvHtbeiq2GwiqJkGS0MiqL4oIVBURQftDAoiuKDFgZFUXzQwqAoig+uLww20nZPnzrJC6OG0a/XbTzQuwub1q91RNdNac7gLr82vMYf2M/AB+6lZ7eO9OrWiU+mfeC3VkRYCJ8PbsicYY34+onGDL6lIgA3VyrMF0Nv5svHGzF90E2UKex/KoGmRHuxlTo86V//oG6DRrz7yRdM+s+nlC5bLmi92khzBnf5teU1NDSMwcOGM/3zObz7wSfMnP4RO3f4p3s+OZW+k36h4/jFdBq/mKZVYqhVugDP316dx6etpdOExcxevZ9H2lTwS19TotNhI2034fQp1q9ZRbtOXQEIDw8nKjpfUHoFO2nO4C6/trwWiYkhrmo1APLmzUvZ8uU5dPCg33qJ5z2DqsJChbBQwWAwxhCV25OXFJ07jPgT5/zSdroNrCQ4iUgE0AvYZ4z5QUTuBG4GNgNveROdAiajZNz169YFpHlg317yFyjI+JdGs3PbVirFVWPgkOHkzhNY8JQNrzZxk98r4XXf3r38tmUz1WvU9FsjROCLoY0oUySSDxfvZu3uE4ycsYF3H6jL2aRUTp9NpvtrS/3SdroNbPUYpgK3AoNF5APgDmA5UA94J7ODgiElOiUlhe2/baFj1zuY+P4McufOw/QPpuSIFyU4SExMYMTfBjP0iZFERUX5rZNqoNOExTR6/mduKJ2fysWiuL9pWfq9s5LGL/zMZyv+ZNRtcQ469x9bmY81jDE1RSQM2AuUMMakiMiHQKZ38oIhJbpI0VhiYmKJq+75zdC4RRtmOFAY3JTmDO7ya9NrclISI4YNoV2HjrRo1cYRzVNnk1m6/SjN4mKIK5GPtbtPADBnzX6m9q/nl6ZbUqJDvJcT0UAkkHZRmQsId+okNtJ2CxUuQpHYWPb8sQuANSuXU7pc+aD0ahM3+bXl1RjDi889Q9ly5bnzrnsD0iqUN4Jo772EXGEhNK5cmO3xp4nOE0bZIp7L1MaVi7Dj4Gm/9N2SEv0usAUIBZ4CPhWRnUAD4BOnTmIrHfiRoSMY99xIkpOSKFaiJMOeej5ovdpIc3abX1te165ZzddzZlOxUmX69vDcjB746BAaNWmWba2YfLl4uXdNQgVCRJi79gA/bz7EUzM2MPHeOqQaw4nEJEZM910aLyu4IiUaQERKABhj9olIATyLz+w2xvySleN12rW7pjGDu/zqtGsPmU27ttVjwBizL932cWCmrXMpiuIsrh7HoCiKHbQwKIrigxYGRVF80MKgKIoPWhgURfHhmkuJtvWYKnd4qOOaxxMdmVLiQ1QuOw+j3LSo7emzyVZ0T1j6N3v8i41WdL/oX1dTohVFyRpaGBRF8UELg6IoPmhhUBTFBy0MiqL4oIVBURQfrE2iulIsXriAcWNfIjUlla6330G//gMC0os/sJ+/Pz2So0cPIwhdbu9Brz53BaXX9KSkpDDg7p7EFC3K2H9ODFjvudGjWDh/HoUKFWbGrC8dcOjBVhvY0D137hyP9L+bpPPnSU5JoUWrtjzw0KCAdf/cvYuxzw6/8Hr/vr3c1W8gXXr0zZZOeKgwpmMc4aFCaIiwZOcxPl69j6LRETzRsjzRucLYcTiRf877neTU7D39d3VhSEvGnfz2VGJjY7mzZ3eat2hJhYoV/dZMSwaOq1qNhIQE7undnfoNGlK+gv+atrymZ+YnH1KmXHkSE/wL+riYTp270qNXH559aoQjemCvDWzpRkRE8NqbU4iMzEtyUhID+91Fg0ZNuL7GDQHplixdljemzrjg/e5ubWnYNPuhKkkphmfmbuVsciqhIoztXIVVf57gthqxzF4fz8KdxxjYuDStqxThm82HsqXt6ksJG+nATicD2/SaxsH4AyxbtICOt93uiB64K83Zlq6IEBmZF4Dk5GSSk5ORjOML/GbtquUUK1GS2GIl/Dr+bHIqAKEhnl4DBmqWiGbx78cA+Om3IzQoWyDbutZ6DCJSHugGlAJSgN+Aj4wxJ506h+10YCeSgdOw6fWNCeN46LHHSUxMcETPFrbawGbbpqSkcH/fO9i7ZzfdevR25LuQnvk/fkvz1v6HsIQIjO9ajeL5cvHVpoPsP3mOhHMppF05HEk4T6HIiOzr+u3oEojIY8CbQG48ydC58BSIZSLS/BLH5XhKdBpOJQPbZsnCeRQoWIgqVavntJWrktDQUN7/+HNmff0TmzasZ+f2bY5pJyUlsXzxfBq38D9kNtXA0M830e+jdVSOyUvJAs4kidnqMfQHanmToScAXxljmovIZOALoHZGBwVDSjTYSQa25XXD2l9ZsnAey5cs5Py5cyQkJPDiM0/y9AvjAtZ2GlttcCUSraOj81Gnbn2WLVlEeQfyJAFWLltEhcpxFCxUOGCthPMprN93irjYvOTNFUqIeIpG4bwRHE08n209m/cY0opOLiAKwBizmyBPiXYyGTg9tpKMBwwaysy5PzJ99neMHvMyderVD8qiAPbawJbusWNHOXXKc+V77uxZVixfShkHlitMY/4P39CsVTu/j8+XO4y8EZ7JexGhwg0l87Hn+FnW7ztFo3IFAWhZuTDLdx3PtratHsM7wAoRWQ40AcYBiEgMcNSpk9hIB3YyGdi2V5u4Kc3Zlu6Rw4d48dlRpKakkmpSadn6Fho1bR6wLsDZM2f4deUyHn3iab81CkaGM6RZOULEc6N08c6jrNx9gj3HzvC3lhXoU/c6dh5J5Puth7OtbTMlujpQFdhgjNmS3eN12rVOu7aJTrv2kNm0a5sp0RsBO38bRVGs4upxDIqi2EELg6IoPmhhUBTFBy0MiqL4oIVBURRfjDGu/wEGXOu6bvLqNl03eXVK92rpMTgXbOBeXTd5dZuum7w6onu1FAZFURxEC4OiKD5cLYXB1hxtN+m6yavbdN3k1RHdoF2iTlGUnONq6TEoiuIgWhgURfHB9YVBRNqJyFYR2S4ijkQai8gUETkoIhuc0PNqlhKRn0Vkk4hsFJHBDunmFpFfRGStV/c5J3S92qEi8quIzHFQc5eIrBeRNSKy0kHdAiIyU0S2iMhmEWnogGYVr8+0n5MiMsQBu4jIUO+/1wYR+VhEAs5kE5HBXr2NAfu0McDiSv0AocAOoDwQAawFqjmg2xSogydLwimvxYE63u1oPOG4TngVIMq7HQ4sBxo45Plx4CNgjoPtsAsoYuG78D7wgHc7Aihg4bt2ACjjgNZ1wO9AHu/rGcC9AWpeD2wAIvHEKfwAVPRXz+09hvrAdmPMTmPMeeAT4LZARY0xC3Awacqrud8Ys9q7fQrYjOcLEqiuMcakLSYR7v0J+I6yiJQEbsWTxhXUiEh+PMX8XQBjzHljzHGHT9MK2GGM+cMhvTAgj4iE4fnPvC9AvarAcmNMojEmGZiPJ6XdL9xeGK4D9qR7/ScO/GezjYiUxROIu9whvVARWQMcBL43xjih+y9gOJDqgFZ6DPCdiKwSEadG/pUDDgFTvZc+74hIXoe00+gFfOyEkDFmL/AKsBvYD5wwxnwXoOwGoImIFBaRSKADnmR2v3B7YXAdIhIFfAYMMQ6tsWGMSTHG1AJKAvVF5PpA9ESkI3DQGLPKCX8X0dgYUwdoDzwiIk0d0AzDc+k3yRhTG0gAHFtCS0QigM7Apw7pFcTTsy0HlADyikj21qe7CGPMZjzZqt8B3wBr8Kzn4hduLwx7+d+qWNK7LygRkXA8RWGaMeZzp/W93eefAf+jhz00AjqLyC48l2ctReTDADWBC78tMcYcBGbhuRwMlD+BP9P1lGbiKRRO0R5YbYyJd0ivNfC7MeaQMSYJ+By4OVBRY8y7xpgbjTFNgWN47mP5hdsLwwqgkoiU81b1XsDsHPaUISIieK6BNxtjJjioGyMiBbzbeYA2QLbDd9NjjBlpjClpjCmLp01/MsYE9BvN6y+viESnbQNt8XSBA8IYcwDYIyJVvLtaAZsC1U1Hbxy6jPCyG2ggIpHe70UrPPecAkJEinr/LI3n/sJH/mq5elFbY0yyiAwCvsVz13iK8YTQBoSIfAw0B4qIyJ/As8aYdwOUbQTcBaz33g8AGGWM+SpA3eLA+yISiqfQzzDGOPZ40WFigVme/wuE4Vmy8BuHtB8Fpnl/QewE7nNC1FvA2gAPOqEHYIxZLiIzgdVAMvArzgyP/kxECgNJwCOB3IDVIdGKovjg9ksJRVEsoIVBURQftDAoiuKDFgZFUXzQwqAoig9aGK5xRKR52uxJEel8qRmq3hmMD/txjr+LyN8C8emkjnJ5tDBcpXjHNWQLY8xsY8zYS3ykAJDtwqC4Dy0MLkNEynozB6Z5cwdmeifNpGUdjBOR1cAdItJWRJaKyGoR+dQ7TyMtw2KL93Pd0mnfKyJveLdjRWSWN+dhrYjcDIwFKnizCV72fu4JEVkhIuvSZ0GIyFMi8puILAKqcBEikl9E/hCREO/rvCKyR0TCRaS/V3OtiHyW9ve76Ph5IlLXu13EO3w7bULZy+k8OTYw6VpCC4M7qQJMNMZUBU7yv7/Fj3gnKf0APA209r5eCTzuDQR5G+gE3AgUy+QcrwHzjTE34Jl3sBHPxKQdxphaxpgnRKQtUAnPfIdawI0i0lREbsQzlLoWnll+9S4WN8acwDPRp5l3V0fg27S5A8aYet5zbwb6ZaNt+uGZrVjPe97+IlIuG8craGFwK3uMMYu92x8CjdO9N937ZwOgGrDYOwT7HqAMEIdnAs824xn2mtnkqJbAJLgwe/NEBp9p6/35Fc/w3jg8haIJMMubDXCSzOevTAd6erd7pfN+vYgsFJH1QB+geibHZ0Rb4G7v33k5UNjrSckGrp4rcQ1z8Tj29K8TvH8KnmyG3uk/KCK1HPQhwP8ZYyZfdI4hWTx+NjBGRArh6b385N3/HtDFGLNWRO7FM2/lYpL56xdb+lg0AR41xnybRQ9KBmiPwZ2Ulr8yDe8EFmXwmWVAIxGpCBeu4SvjmXlZVkQqeD/XO4NjAX4EBnqPDRVPStIpPLF0aXwL3J/u3sV13hl+C4AuIpLHO5uyU0Yn8CZPrQBexRMfl5YfEA3s905T75OJv114iglA94s8DfQei4hUFudDW656tDC4k614Qk42AwXxdvnTY4w5BNwLfCwi64ClQJwx5iyetQ3nem8+HszkHIOBFt7u/Co8+ZRH8FyabBCRl72pQx8BS72fmwlEeyPspuPJ4Pwaz3/+zJgO9OWvywiAZ/BcBiwm8ynkr+ApAL8CRdLtfwfPlOvV4gnznYz2jLONzq50GeKJhZtjjAkopUlRLoX2GBRF8UF7DIqi+KA9BkVRfNDCoCiKD1oYFEXxQQuDoig+aGFQFMWH/weGtiz7HJf4dQAAAABJRU5ErkJggg==", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "from sklearn.metrics import confusion_matrix\n", "\n", "mat = confusion_matrix(ytest, y_model)\n", "\n", "sns.heatmap(mat, square=True, annot=True, cbar=False, cmap='Blues')\n", "plt.xlabel('predicted value')\n", "plt.ylabel('true value');" ] }, { "cell_type": "markdown", "metadata": { "deletable": true, "editable": true }, "source": [ "This shows us where the mislabeled points tend to be: for example, many of the twos here are misclassified as either ones or eights.\n", "\n", "Another way to gain intuition into the characteristics of the model is to plot the inputs again, with their predicted labels.\n", "We'll use green for correct labels and red for incorrect labels; see the following figure:" ] }, { "cell_type": "code", "execution_count": 32, "metadata": { "collapsed": false, "deletable": true, "editable": true, "jupyter": { "outputs_hidden": false } }, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "fig, axes = plt.subplots(10, 10, figsize=(8, 8),\n", " subplot_kw={'xticks':[], 'yticks':[]},\n", " gridspec_kw=dict(hspace=0.1, wspace=0.1))\n", "\n", "test_images = Xtest.reshape(-1, 8, 8)\n", "\n", "for i, ax in enumerate(axes.flat):\n", " ax.imshow(test_images[i], cmap='binary', interpolation='nearest')\n", " ax.text(0.05, 0.05, str(y_model[i]),\n", " transform=ax.transAxes,\n", " color='green' if (ytest[i] == y_model[i]) else 'red')" ] }, { "cell_type": "markdown", "metadata": { "deletable": true, "editable": true }, "source": [ "Examining this subset of the data can give us some insight into where the algorithm might be not performing optimally.\n", "To go beyond our 83% classification success rate, we might switch to a more sophisticated algorithm such as support vector machines (see [In-Depth: Support Vector Machines](05.07-Support-Vector-Machines.ipynb)), random forests (see [In-Depth: Decision Trees and Random Forests](05.08-Random-Forests.ipynb)), or another classification approach." ] }, { "cell_type": "markdown", "metadata": { "deletable": true, "editable": true }, "source": [ "## Summary" ] }, { "cell_type": "markdown", "metadata": { "deletable": true, "editable": true }, "source": [ "In this chapter we covered the essential features of the Scikit-Learn data representation and the Estimator API.\n", "Regardless of the type of estimator used, the same import/instantiate/fit/predict pattern holds.\n", "Armed with this information about the Estimator API, you can explore the Scikit-Learn documentation and begin trying out various models on your data.\n", "\n", "In the next chapter, we will explore perhaps the most important topic in machine learning: how to select and validate your model." ] } ], "metadata": { "anaconda-cloud": {}, "jupytext": { "formats": "ipynb,md" }, "kernelspec": { "display_name": "Python 3 (ipykernel)", "language": "python", "name": "python3" }, "language_info": { "codemirror_mode": { "name": "ipython", "version": 3 }, "file_extension": ".py", "mimetype": "text/x-python", "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", "version": "3.9.2" } }, "nbformat": 4, "nbformat_minor": 4 }