{"metadata":{"kernelspec":{"language":"python","display_name":"Python 3","name":"python3"},"language_info":{"pygments_lexer":"ipython3","nbconvert_exporter":"python","version":"3.6.4","file_extension":".py","codemirror_mode":{"name":"ipython","version":3},"name":"python","mimetype":"text/x-python"}},"nbformat_minor":4,"nbformat":4,"cells":[{"cell_type":"code","source":"# import sframe                                                  \nimport numpy as np         \nimport pandas as pd\nimport matplotlib.pyplot as plt\nimport seaborn as sns\nfrom scipy.stats import multivariate_normal                    \nimport copy                                                  \nfrom PIL import Image\nfrom io import BytesIO","metadata":{"_uuid":"8f2839f25d086af736a60e9eeb907d3b93b6e0e5","_cell_guid":"b1076dfc-b9ad-4769-8c92-a6c4dae69d19","execution":{"iopub.status.busy":"2022-07-17T12:41:20.350869Z","iopub.execute_input":"2022-07-17T12:41:20.351584Z","iopub.status.idle":"2022-07-17T12:41:21.845893Z","shell.execute_reply.started":"2022-07-17T12:41:20.351417Z","shell.execute_reply":"2022-07-17T12:41:21.844425Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"TLDR:\n<br />\n![image.png](attachment:b93e3550-be57-407e-99de-d4acdffba088.png)\n<br />\nStep 0:\n<br />\n![image.png](attachment:eec8460b-d0f2-4bd0-9c33-55c41db1c79f.png)\n<br />\nIterate between E-Step and M-Step\n<br />\n![image.png](attachment:ffafe0e1-e414-4757-907a-da9b8843dd1f.png)\n<br />\nAn Alternative Generative View\n<br />\n![image.png](attachment:82a2752a-d091-4098-b381-a9ea67c4bf83.png)\n<br />","metadata":{},"attachments":{"b93e3550-be57-407e-99de-d4acdffba088.png":{"image/png":"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"},"ffafe0e1-e414-4757-907a-da9b8843dd1f.png":{"image/png":"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"},"eec8460b-d0f2-4bd0-9c33-55c41db1c79f.png":{"image/png":"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"},"82a2752a-d091-4098-b381-a9ea67c4bf83.png":{"image/png":"iVBORw0KGgoAAAANSUhEUgAAAqMAAAHrCAYAAAD7fJT0AAAgAElEQVR4nOy9fXAb553n+eWLRUuy0XA8UqSLoY7WE0sKOc1JZnVkmCMmsTQmmVmoxpwZwndHXOR1EdnbwOJkj9Cei5T+kIhylcCthFJ76wyeV8qCmRLoOdknXEJSY8k58IKIp0wy7AVtyU6stKAZceVzgqcTSqbEl/uj8TTxylcQ4MvvU6UiATS6HzRA9Rff31vR/XtjMyAIgiAIgiCIAlBc6AUQBEEQBEEQGxcSowRBEARBEETBIDFKEARBEARBFAwSowRBEARBEETBIDFKEARBEARBFAwSowRBEARBEETBIDFKEARBEARBFAwSowRBEARBEETBIDFKEARBEARBFAwSowRBEARBEETBIDFKEARBEARBFAwSowRBEARBEETBIDFKEARBEARBFAwSowRBEARBEETBIDFKEARBEARBFAwSowRBEARBEETBIDFKEARBEARBFAwSowRBEARBEETBIDFKEARBEARBFAwSowRBEARBEETBIDFKEARBEARBFAwSowRBEARBEETBIDFKEARBEARBFAwSowRBEARBEETBIDFKEARBEARBFAwSowRBEARBEETBIDFKEARBEARBFAwSowRBEARBEETBIDFKEARBEARBFAwSowRBEARBEETBIDFKEARBEARBFAwSowRBEARBEETBIDFKEARBEARBFAwSowRBEARBEETBIDFKEARBEARBFAwSowRBEARBEETBIDFKEARBEARBFAwSowRBEARBEETBIDFKEARBEARBFAwSowRBEARBEETBIDFKEARBrCkY0+DvDRR6GQRB5AgSowRBEMSaosXZChbTCr0MgiByRNH9e2MzhV4EQRAEQSwEVY1i7779uP7+NYiipdDLIQgiB5AzShAEQRiEQmFUVR/Ejp17sHff/lUXDmdMd0TNZqHAKyEIIleQGCUIgiAAAJ2eLtTVN6JSKselwQu4/v41sJgGVY0WemkGsRgr9BIIgsgxpYVeAEEQBFF4ZLkHHk8XfL5uOJrtxv0uV0sBV5UdVY1CksoLvQyCIHIAiVGCIDYEoVAYoaEwbqlRCGYBgmCCo9lOeYfQQ9/yaz44mu1JQnQ9EgqFUVlZAUEwFXopBEHEITFKEMS6p8n+IoLBfgiCALPZhFhMA2MMHk8X2tvb0NHeVuglFpSRkQhUNQpXwFnopSyYbOH6UCgMUbRk/ZLRZH8RlVI5BgcvrOTyCIJYBJQzShDEuoYxDcFgP4avXsbYnRu4/v41jN25geGrl2GtrYHH04VOT1ehl1lQeFHQWnALeeGSoowm3c+Yhib7i6irb8TeffsRDA5keb4JoaHwqivMIoiNDIlRgiDWNYJggiTpYVl/b8AoxpHi7pgkVcDj6coqXjYCPPcyNBQu8ErmhzueMZbsjJ6RfQgG+yFJFQCQsehKlnuMx93u4yu8UoIgFgqJUYIg1j3DV9+B2SxAlnvSHLEeXzcAQJZ9hVjaqkAULbDZGuB2H0cotLoFKXdvU51RWe6BIAhgjEEULXA4knNfFWUU8ms+dLS3wdFsB2OM3FGCWCWQGCUIYkOQLQQtSeWw2RoQGgob4eqNSI+vG5VSOerqGwuetsCYhlAoDFnuyehwWmtrkkSzooyCMQbGGGIxDX2Bc0nvt6KM4rm6Rri+7YQklcPl0nNjN7IbThCrCRKjBEFsKFIdNWA2TD0yEsnLGurqGjOuo5AIggmDgxfQ3t4Gj6cLVdUHC9JflDENLc5WNNlfRKenC3v37U9za10uZ5KzyYuZJKkCw1ffSWr5JMs9eK6uEYds9UabKh7qV5T8vN8EQcwNiVGCIDYMZsEElqEKuzKeR5gvgShVVqzaHpkd7W0YvnoZjDHs3bcfstyTtg1jGtxHj2PHzj2oq2vM2bEZ03BG9iEUCqPH142xOzdgszXA+a3WpO1stnpYa2uMvM/KSv39MwsmQ2jyNbqPHkOlVA6v96TxfC6yBYGmOBHEaoDEKEEQGwZRtGAkg+DMVBTDWObJQ1zkZHLsFor31IklPU9RRg0RuGPnnhVzLiWpHNffvwaXywn30WNJgpM7l35/AC5XC2IZzlMoFE66jzHNELWqGs0ocAHA7w/A4+mCw2GHzVYPQE8fUNVo2jH6+s5BFC3o9HRBiIvQ0FAYO3buQVX1QezZux+yrPdOrbXWYM9e/f0KhcJoceritnKVfiEgiI0GiVGCIDYMglkvcEllJB6uTexNeUb2oa4+2fVLFGKSVIHg/zV3ziFjGnbs3AP30eTK7aXkKirKKJrsh+H3B+D1noDZvPJtmLynTmD46mWMKKOoqj4IYLZqvSPen9V76gScztak1+T8VmvSa/b7A3AfPYZQKGy4lXMJeWttjfG7IJggCEJapb8gmDB89R2jSX9f4BxstgYwxqAoEYiiBd5TJ+H1njQKlurqG1FX3whFiUCSKpLcUoIgCgeJUYIgNgy8qCXVZeO3zcJsD0uPpws2W0PSdn5/IEmIKQk5prLck+b4CYIJDocdsuwzHmNMy1gg1GR/0SigylRI5XYfQyym4dLgBbCYlvMQM2Na1nzaS4MXoCgR+HsD6O3VhTjPv2RMQ2ho1gnljnLibZ7b6T56HMFgP4DMKRHcmU7N+WSMQdyVuYk9/wIhSeXoC5zF/XtjuP7+NQxffQcuV4vhmvKiJUEQ4HI50Rc4uyb6qhLERoDEKEEQ6x5VjUJRRo3c0EQxysPGgiDAatUdOe7aJTp0XERaa2vgcrXgYrDfqMDnjylKJO5gvmg8j++DF8uMjETS3NlQKIxgsB+xGEMwOIA9e/cbTiSgO6mhoTA62tswokTgPnoMqhpFVfXBnHUAOCP7slbRS1I5JKkCgcBbUNWocZ5UNWqEvLlzyc8dP8dnZJ/x2rkjCQDya7408c5vy6/1IBgcQF1dI9xHj8HlchrHXAiZpi95T53A/XtjGLtzA95TJ2gMLEGsIkiMEgSx7gkNhfFcXaMhaLgrx/M/GWOGi6Y/rounROcsFAqDMYZmhx2MaRiKj50UBJPxmN6r8xiCwf54uyHN6F/KXVbGNMRiyQJSvaULNy7uKqVyw4kEYLiJADAUF3uHbPV6Rbk/N70yFWUUihJJc435OVKUCP74j/9Iz82Mt12qq280epSGQmEoyig6PV0QRYseFq9rNBxm76mTaG9vw6V4xb4kVSQJQt6eCdB7vjbZD0O9FdVD7UvMsSUIYm1As+kJglj3HLI1wOlsxRnZB1G0IMaYLrLiwlGSKvCya3Yuu2BOD4FzkSbusqDF2YrQkC5GE91Bf2/AcAirqg9AkiqgKBFYa2uSnL1MeasAUFffCEmqQF/fOVRVH4DH0wVxlwW7d4sAAPfRYxAEAT5fNxzNdozExR8PmS8H17dbjFGagiAYgjIW08AYgyRVwN32MnZ8djvcR4/BfVR/XT5fNxjTUBU8gKrqAxAEAX2Bs+jtDcDfG4C1tgY9vu4kYd/R3pZ2fN6eSRQt8L3eDbNZMMQ+QRDrm6L798ZmCr0IgiCIlWbvvv1Q1ShOnTqBxx97DD/q/3sEg/0QRQuGr15OEj2dni54PF0YHLhgiMhgcABN9sPGNlxo8txNLjAdzXY0N9vh8XQhxjRYrTXoaG9LmhxUVX0gad+qGsXefft1IRo4C1G0GAVLqhqFIAiw2erhaLajsrLC2Bdf5/X3r+Uk7Kwoo/D3Bgy31Ww2QdxlQXt7W9Jx9derpTWWvxjsxyFbw5LaVoVCYcNpvf7+tWW/FoIg1g4kRgmC2BD4ewNwOpP7VSaKv0RUNQr5tZ4kEQnogpTnTIqiJZ4POQpHs113XGMsTbRlQpZ74uHt+qT7UwUeX4vZLKSlDLiPHoeqRsEYg/fUyZy4o4WEi1FBEDB250ahl0MQRB4hMUoQxIZBlnv0QiGmwdFsh8NhX1NhYJ6TGQz2GyFyI7QfOFvo5S0LvQ3WMyRGCWIDQmKUIAhiDcDnqwN6ziV3Qp3OVlwMDqx5AcfFKADcvzdW4NUQBJFPqICJIAhilcOYZuSr9vi6k8L7u+KFRmsd3tyexYvL1pJjTRDE8qDWTgRBEKuci8F+qGoUHe1taXmmvHdqrvqNFhI+nnOlxpwSBLE6ITFKEASxymHxvqSZqtSzTZVai0iVurDmbZ4IgtgYkBglCIJY5Qh5mEO/GjCE9a21L6wJglg4JEYJgiBWOeZ4L9NM89wVZdRoUr/W4SkHBEFsLKianiAIYg2wd99+xGJaUtU8Y5ox6Wmtt3bidHq6jL6tBEFsDEiMEgRBrAF4035rbQ1cLidijOlTnmIaLg1eWNLUI4LIBC+Go44GRL4gMUoQBLFGkOUeuI8eM25zYZpaYU8QyyEYHECLsxXDV98hh5rIC5QzShAEsUZwuVpw/94Y+gLnMHz1MgYHL5AQXeM02V/MeyeE+Y5ntdaAMYZgcCBPKyI2OuSMEgRBEESB2LxlBwYHLsBqrcnL8UKhMJrsL847sWvHzj0QRQuGr76Tl3URGxtyRgmCIAhiHhjT4O8NwN8byJljyB3K5bSyqouPiF3wMTMcS1FG09zSSql8XfSuJdYGJEYJgiCIVQcXf4WGMQ3uo8exZ+9+OJ2tcDpb0WQ/jL379i976lUunh8aCi96P3zkKsftPgazWUjaRqqsAGOMBCmRF0iMEgRBEKuOWIzB6Wwt6JhTVY2iqvoA/P4AHA67nqc7cAGOZjtUNQq/PzdimU/YWiqhUHjRz0kUmTGmpVXOr6fJXsTqh8QoQRAEsWrJJIYY0/IiUoPBAahqFJcGL8B76gQkqRxWaw283pMAgNDQ4kVgIssdeyoIJgiCgBElsqxje0+dSHucV9FnGrRAELmGxChBEMQapvjDMEp/2IXSH3YB9wvnIq4UmQSb3x/AGdm34sfmgiw1XYALZLZMMckFdYwtfT+iaMkqGBnT0OnpSrqPT/NKFPOZiqf4dgSRD0iMEgRBrFFK3u3Bpu81ovRHXSj9URfKXj2wuB2sYvHKcxgzCa3QUDgv/S95qFrclXysi8F+AEDtMivglyNCOZVSedYwvSCYMBQKZyy4mu/Y/LXnYo0EMR8kRgmCINYgJe/24JG/O4aZJy2YeOUyPu36AEWfRFF0e2Fh1aJPoig7th+P+FtXeKXzoyijqKo+mFF4ZhJDihJZdp7lQqisrACQHI5XlFHIcg8A4JCtIWfH8vcG0OnpWnT+p2AW0gqSEunrO5dxktJ854+L/VspaRIUtidWAhKjBEEQOcZ99PiK7r/okyhKf9iFmc0CHjjPYeapcmCzCTObBWDzwkY4llwNoOg+WxXuqKpGYRZMSSNNeT5kIcWPIJggSRUIhfSKdb1H52EwxuByOXM2grU3PurV4+lCXX1jxnZNoVAY/t5Amlidr9BIEExJYfiFOJ6qGsWbb74NAPhR/9+jqvog9u7bj5+Eh9FkP5x2rHzl8BLrFxKjBEFsWBjToCijCIXCOasaZkyDLPtW9OJc+qMuFN1nmPp6iy5EF8t9DSXDhW+bxLFaazCSLe8xnpeZeD5jMS1v4WM+jajF2WpMS3I029HR3rbsffO8TFWNwuVyoi9wThe/Q+Ek0el0tqKuvtH4mRh25w4mL2JiTEMwOLCstliJ4vLRR8vQ0d6GvsA5fPTRrxGLaUkpEqFQGHv27seevfup8p5YMiRGCYLYkDCmoanpMKqqD6CuvhF79+1HXV3jsp24Fb8g39dQPDKAmSctmHzWmfTQw2+dxcyW+QtPSq/4UPSJvs6ie4XPCRQEE3p83Wn3i6IF/zgSQaenC01Nh5NEWL4cU+4kBoP9hiPq9Z5MC337ewOoqj6YNN5zvi8kfB82WwO8p07AZquHy9UCYDY1gDfat9bW4Pr712CzNSAYz1kFEgqSYlo83eEAmuyH4XS2QlWjaWvgubip4fdEJKkcJ0+2AwBKS0ths9VDksphFvSUAFnugaKMwn30OJrsL8ZfK40PJZYOiVGCIDYkF4P9CA2F4Wi2o729DTZbA0JDYVRVH1iWoOQOVaY8vVzAw+tTVfa0kPz0F2rmD9Pf11Dybo8e0l9F2Gz1SbcVZRSffPIJNE3DLTUKn6/b2CZT8ZL76HE4nbnPf62U9LxRa20Nhq9ehvfUiYzvrcfTBVWNwlpbYwg+vz+AzVt2oKr6YEankuekJn7eeC4nf40eTxcEQYDP1w1RtMAsmLAr4fUbYfpbUbQ4WxGLaXC59C8poaEwzsi+tIp6QO8tysk2xUkQBKhq1HjcZquHo9kO99FjqKo+AFn2oVIqx6XBC8bxCGIpkBglCGJDwi/67e1t8TDkWQwO6BfV5YgaRRmFsIJtcYo/1C/4009VLOn5pT/UQ/yTfx4PM+c4Z5QX98wFn2qULbeW52aWlZUB0N+jRAFqFkxJYWxZ7oEs+wwXMZfwvFCpsiJrjqj76HGoahSiaIHL1WIIRMGs/1SUiB5izyL6VDVqCFLuevIKfr5fUbQYE5dkucfYngtfWfZBUSLoaG+DtXY2R5SnonAyCfls40j5tkLCdCafrxvDVy+jL3BOHwIweCEvnQ2I9Q2JUYIgNiT8AproSlmtNUbOHkdRRhcVflTVKMzmlXFFAaD4tu68Tj+T0lbovoai26MoHhlAybs9hmhNeu6HYZS+68Pk152YlurTHs8F8mu+NDGfGio+I/sgy76sleOVlRXoC5zDN77xHID01IfEUZXB4AA6PV2GAOvN0VQkjtFiamT+xvLmBMeUMQ0eTxdE0YK+wDmIoiVtdKcgmOBotoMxBvfR45DlHoSGwrDW1sRzVfVtVTUKp7PVcO359v7eQJIQtNbWwOVqMdx5a20NhBThzgmFwuj0dIExLWM4H0BcWAtwfbsl6X5JKjdC9/x18HUSxFIgMUoQxIaEX0id30oWTvzizi/OomhZ1IQbVY2uqDMKADNPWoxwfMm7Pdj06kGUHduPslcPYJPvMB75u2PY9L1GlFxNEGb3Nb0C/0kLJv+8zcgtLbqf25xRm60B/t6A4capahR79u43cguB2XzPbEVAQryynouc1PxQfr/8Wg9a4sLX6z1pCL5ciqLZKUfZc1SNfqQJwnBkJAJVjcJma4AoWhCLaZCkirQQf3v8HASD/XAfPQZBEIz7BMEEXzyX1t8bQCymGeH6YLAfbvdxCIIJtoQWU8HgAGS5x3BTX3Y5USmVG+dQEExxl3X2fc/2eXU023Hj+rWMTfHTz4FAYpRYMiRGCYLYkHCx6fq2M+m+UCicJBoEwbSoymnGWJJDlmt44RFn5qlyzDxpwfQzNZiqtmOq2o6Hjm5MvHIZU9V2Y7uSqwEUfxjGw788mZxXei+3YXp+rrhIDAYH4sUt/YYDx526xLB3VfXBNDHD8zVTK+dfjudE6l0LGFyulrhbp4uyXOeOVkrlYIxld3Lj69yVIVwdDPbjubpGMMbS8mIBXcAOX70MR7MdNlsD+gJnk8Sfo9mOsTs3MHbnA4zduQFHsx3eU/o4Ul7s1Bc4awjxFmcrGGPGNoJgwuDghaRz3Rc4B5+vGy+7nBAEE8bu3Mia47yY3GfGmBFJIGFKLAYSowRBbEj4mMkYY+iMF580NR3OKhoSYUzLmpuoqtGkHDtOTiuNEwTk9Bdq8NB5Vv/3Vycx+XUnpqrtyS2f4q7o5NedmK6Mv7YF9iNdLNypU5QIgsGBpJQHT7ypO2MMklSRlgeaek65KBvK0FvT5+vWQ8gupyFOuRAODYWxd99+VFUfRFX1wUU3kk9FihcaZavgt9nqDXGXuHZrbY0hyrynTmb9UiNJ5fD5utOEaCKJotBmq8fYnQ+S9peYJ9oXODfnZ1iSyuFoti+7yE5RRlFX14gdO/cYTmtV9QG0OFupmIlYFCRGCYLY0MhyDzyeLuzdt9/I1+OiIhQKpwkQxjQ8V9cIj6cLVdUHk3LtIpH3ASDNGfX3BtBkP5wTQTrzpAVF91n6pKX7Gh7xt6Ls1QMoHkk+TsnVAGaeKp8tWuL72izkPEwPzFbGd3q6oMRTHETRAn9vwCjQSRVdeh5oskvLhW1qriUw6xgmVrcLggmDAxeMkDEv/lluz1dHs31B26SKu8HBC7h/bwxjd24YLmauSD2Wz9dtuKfzfZnKBfzvIDQURqVUboT6r79/zXBwCWKhkBglCGJDwgUKd3QEQYCj2Z40PtH5rda0cKPbfQyKEoHv9W4MX33H2JYxDW3ujrTjqGoUbrdeNZ6LiT0P/1IPv276XiNKf9iF4g/DKB4ZQNmrB1Ayogu91Ib2M5/R80SzuqE5rqh3NNshihYoip43KUkVRtiYu5+pYsX2r+qT3D1OX+Ashq9eXrCLZ7XWYOzODUMU9QXOLlucSVI5xu58kHNBmWtWqp1YJnghlc3WkFRRT5OYiKVQWugFEARBFBKbrQGOZjus1pq0i3kspqWJ0dCQnlOa6ux1errws5/9I4DkVjjuo8eNi3YuWuBMV9Zj8utOlL7rQ+mPuoAfzT7G709tZG+E5qEXPJW+6wMAFE2MAwAe7fgyZopKjG0evPKOXiS1DHyvd6OuXm9lJIoWo0elvzcAR7M9TZjPVSSzFBGf63ZD+RR6awF+frONJyWIxUBilCCIDY1e8bww54yPDXW5GpLu530uD3/zf8C57/9t0v3BYD8ee+wxeE+dyNmaJ//qBKaq7Si+HTEKmqYqGzDzVLkuRlND+AlMf6EGk3GHtCh2ByU/exsP/6IDKNtiNMJfrhAFdHHpPXUS/t6A0RrI5+tO6xlKrE14Wyp/b8DIuQZy/yWA2BiQGCUIgshCpVSO0FDYCM/yEGSi+8MvxjZbA77xjedw7vt/C1WNQlFGjck3n/3stpxfpGeeKsdUhrn0M09adIF6X8sYlk993mT93+R0XYm4XC1poW0SK+sHl8sJf28Anvjn3OVyzvMMgsgMiVGCIDY0jGUv4PH5uo0LLTAbSpblHn2yzUgEoaEwRNGSNFs9GOyHokTAGMM3v/nf4+23f4Sq6oMYvvrOyr2QOFNVdpS827Ni1fIEwZGkcgxfvaw3399lgcNBRUvE0ii6f29sptCLIAiCyDehUBh19Y0QBAFjd24s+Hn6pByfcdtaW2M0Igf03FEuYK21NUZBlCz3wOFYfjsdgiCI9QaJUYIgNiSMadix85lFi1FA76/I2wZlKq5hTMPISASVlekTd4i1g6KM5qQDAkEQc0NilCA2GjOTwMwEMP1A/x2TwPQEMDOl3z8zGd9uavZ3fjsRXn1dVKr/XlQKgP9ekrJthoygohJ9++IyoKQwgm3vvv0A9N6IBJFKVfVBY7oRQRArB+WMEsRqJlEYYjIuEKcSHptM2X4q83O5yOSiMydrm0r+uVyKy4DSJ4DSz+RNnA4OXKC+iMScKMooiVGCWGFIjBILg7tpU/fiP7X4fQlCpLgMhjNWXAYUlQElW4DirQVb9qrDOGeTczuT3JXMldBbC0xPAA/G9H9FJUDZ54FHtq3oIUlkEPNBM9YJYuUhMUokk010Tt2b/7nTE9kfK/0MULwpr67XipHNrUx0KjOFvXPpSq53ZqaAT38FTI/ropQgCoBZMCW19soljOkDFWIxBsY0I7dYFC30JYnYcJAYXUskumqZBNCcpL7VKe7b9PjKiqXJ3+g/H4zFl/MZoPRx4JGdK3O8uZjLnQQSQuGT6fmTJCbzy4MxYPK3wOZnyGFfI7CJInT+9DH439uMym0PMfjXvy30kuZEUUYhCKaMAlAwC2nOqCz36ONgE1p5LRTGNHR6uuD3B+ZsKSYIArzeE0ue784L6BRlFOotff0sph9PkipW/VhTYuNBYnS1kVRc8qkukKbvxX/O4TyuNSZ/o//7VAU27cid+2WEuPn5mpw9h7nOmSTyw/QEcO89YMsXSZCuclStBFW9T4JNFEE0TSF0exM6f/oYOr7y+0IvLSt8MEFf4GzaY6JoSRt3yfvKLoWRkQhk2QfvqZNGlf7dj/8/PP7YVmzevNlwS+XXfHA6W1EpVSypmv+M7Evqj5tMAOqtaE4nghHEciExWigMR/LerOic0jamWHowpr/2R59emNjIJNinNHIv8wTTxqFG70Iq352/g85MxQWpFM9NJlYjzkET2EQRXF+6h46v/B5NF81gE0VZt090UQHAXDYN15fvwfWlBaQF5YgeXzeeq2vM2MZJEExgLDmMzmIMWKIYtVprjHGofIBCk/1FAMliWL0VhSz7EItld0/nPE5tDSoD59BkP4z//P3/DR999Gt88Yt7EWMMLKZBMK/xVCli3UFidLmkhXw/jd8/lRLiTQj7kmhKZ+pesvuVTXBuRLG+ynB3vIGYNo6+c6/k98AzU8D994CtX8rvcdchdXWNaG9vMwRRLvC/txmh25sgmqbQ8ZXfQyjTuwZaLQ/mfY7ji/fBJooQur0J7h8/DtE0BdvT+YkECYIJw1ffydhVwSwIAPQiJkkqRzA4gJu/ViFVViz5eB3tbWn38RA6Yxrc7mPw9wYgSRWoXOJx+PsqihZcuRLCW2//EDeuX6Oet8SqhcToYpke1wXSFCNxlGtmpoDx/6JXUtM5XbXEtPHCHXx6Qi9sevTpwq1hHTCijEJRRnMqRntHHwUA+Oo0Q4iqWgmkbXPntDvK7xtOqKqVYO8bf4DOnz6WNzHKySTUeDj+H37+j+j0dOkjXmNaTkUdizF88pvfIhQKw/mtVqhq1JjqtdzjiLssuPazXwDI/PoIYrVQXOgFrHpmJoGHd/QL4D1FF0v3b8RDy/dINK0EdE5XNWr0bmEX8PDj9ZU/vQ5QPi5F6PYmSNsmYX1q1gkVymYgmrL/PTu+eD/pcdE0BaFsBsrHpXOG91NZqV6x4+O6SG5r64CqRvGf3ngN2u9+l7P98xzR69c/QF19I2IxDd5TJ9HXdy4nFfWiaMEvf/lRDlZKECsLidFMzEzqruf9D4DxX+hFNg8/Xlh7I4JYpdQ935GT/ZhNWwsvSB/cLuzx1ziiaAOeslEAACAASURBVEFoKDz/hgskdHsTAKS5mX2HYkm3lY9LIf9iCzp/+hgAXaymPsdcNm08tlBanK3YvGUHFGV0zu14NftCxGsoFEanxwsA+JMv/zEuDV7A5z6X2+4fZ2Qf1FtRTE9Pw1pbg0uDF+BytWR0MVU1iib7i2iyvwinsxWy3DPv/gWzgImJCSPvlSBWKyRGE5nSgIlf6wL03nt6tTe5dMQ6YSRyM2f7Um+RO7qWsVpr0qrEl0MoqovRyu0P9du3N0H+xRYEf1UG948fR92bT2DvG3+Aqt4n0fnTx6B8nDlDTNVKoGolc7qpGZ8Xb7/kdh+bczteZX5G9s25nSz3oMn+oiEKrX/61TnD3MHgAPbu2w9/b2BR637Z5cTX/vS/w8zMDPr6zs1ZOS+KFkhSOUKhMPy9gQUdy1o7m4bBz5GijMLpbF2wKCeIfLCxxejMpN7D8MFtPQR/7z09/E4ClEhBGc2dkFsuavQu9v5LJ0LhyKKfy3KQ7ylatudkP8uG3NElw6vEczVdiItLadskgr8qQ92bT8D948fh/vHjkH+xBaHbmxCbKEZ79TiGmz9Jc0w58s+3AACav/jpoo7vaLZDEASEhsLGawoGB9IEN3dOK6X5C4O83hP42x/870n3ZRJvijKKFqee65m9nVJmBMGEij/6IoC5Jz1xV1SWe4z+pGbBBPfR4/Pun3Mx2G/sa0QZhcfThT1798/rJhNEPthYBUwzk8DU73THc3qcwu7EgvAHrsDjPQ8A8J0+AmvN0itpc0GwfxjWmoolrSMXLZkEYauxL9GyfVn7WhaTvwGw8oVMXICoatT4J4oW2Gz1K37sXMIboQOzYiw0FIZDXFpjdY7ycalRqCSapiCapuD60j1I2/XCpcvqJgSuP4r6z0/M2W80dHsT/O9thlA2g5e/vLgvOy5Xi9EOKRgcgNVagyb7YQBAX+AcbLZ6MKYZ4jTRgZTlHghmU1KDed4UngvEbA4iYxpanK2GQIzFFu80csE4VxsnVY0iGBeTgiBAFC0QzALEXXPnlVqtNbDW1uD2P/0zZLkHjmY7bLZ62Gz1UJRRPFfXiOfqGjF258ai100QuWRjiNGZSeDhGIX2CoQyehOCaWthhcsyCPYPw9ZQhQ73C3ju+Q70nXuloK9FWWK4Xdy1HaGfRHIuRtXoXajRu/kX6TNTemSj9Ill7YYxDX6/3ghcGYkgxjQwxgxhkWlSjiRV5FWMKsooLgb7wZiGjva2RVdGq2oU7qPHDUFjMpmM+5dLbEIPsPFcTwDwfk0v8lG1Erh//DgA4OqdR7Lug00Uwf3jx8EmitB3KLaofFGOtbYGsuxDp6cLHZhtn9TibMWN69cwMhIBYwySVJFUHCS/5oPr286M+zSbhaTb/HmKMmoIUUWJwFpbgxjToCgR48vKQuHb8klJGV+btQb3740l9TtdKIODFwDo/UwTv3xIUjm83hNwOlsRCoVz2lmBIBbL+g7Tz0zO5oBO3F6UEGXa+JLCoPlC9gWx4wv/Y+ELSRaA/HoQVc9+J2uomzdRX60okZsQLdshmLZi+PJ3Cy6ql9payWzauuDnqtG7cB45DdkXTHusMi5muShWIjfR9M1Xl7SmZXP/hh7lWCbBYD9CoTDUW1H8+te3oKrReNGHLkQff/zxuABtQHt7W8ZpPSsFd7A8ni7Isg87dj5jFK/wgpzEcHRq2JUxDU5nK4LBfjia7fD5uvHEEwJMjz+O//PtHy57fVyEqlpJ8ro/LkXdm08YVfFctGaC55G6vnRvyS2drNYaCIIAxpghuiWpAowxuN3HEkL06XmZijL3//W8D6ggmPR0gFAYLfFzKooW9PWdM9xUp7N1UevmvUzZAlzV5bRn6gucTRsvyp3VuYQwQeSD9S1GJ9Ql5YD6A1ew50+cqHu+Y9UKvlA4AutXK2AWVv94REHYCrOwFTGWLhqYNo7nnu/IKHpWC6vx/V8KUsXuBbuqomW7sX1qfih3RrmwtTVUoe/7eW6An8j4f9GLD5eIIJgwOHgBw1ffwfX3r+HNvnMAgLE7H+D+vTFIUgWOHzuK4avvoC9wFh3xCTr5gDEtJdzcAACG2OJ5ivJrs+L0ubrGJEF0RvYhNBSGzdYAn68bqhpFc7Mdn/vcToy+d33ZRSw8PM9dUD4C9Lk3P2OE74WyGbCJooztmvzvbYb8iy2Qtk0ua2yoIJhwKO5W804B/L1KLPipTXEA29vbcDE4kPE8cPEnJDikPb5uQ/AKggDf63o/UEezHS6XE6Gh8KLyMK3WGthsDUsa+7lceuPnhMU0+HsDcB89ntPCNoJYKOtXjH76Kz0sv0jU6F24O94A08Zha6gC08bhD1xZgQUuDxYXdoJpbYjRGBvPGMYN/SQCJXKz4HmYGwFB2LqoEL/LaYPv9JG0zxh3hhP3VfD3L0cOaSI8v5IxhliGUH0+OCP7oKpRuFwtSWkBfAIQL0rhzp3fH4ivd1ZYccHhaLbDffS4EeqfePAAMzMzSx45mUj7V/RzL/9iC+refAKeq1uTxoJm6xmqfFxqhPGtTz3AmZ9vhXPQZPzjLaMWSnOK8ydJ5fC93q0fS4lAEISkCnNAPy9jd25kdR1F0ZL0mM1Wj+Grl+HzdePG9WtJ4W3vqRMYu/PBooSlIJjQFzib1zA5YxrcR48bAt199JjhnmcS5Xz7HTv3UAU+sSKsz5zRB7eXJEQBwOM9D6aNw+W0wXvyJdQ934FbhW5jk4GYNm44VKudyvLdRig+NcQd7B8GoLt2QNzxLbSwyUBBpw6lwNg4zEv4EiJatmd0pxcLd+Pfe/8Wmg6/CsbGIVXshvfkS2nbhsL6lw2pYvfKvq8zU3pf4ByMCuUjGENDeh5dLKZldbqMvpUxBp+ve9nHTkVRRiEIAl526TmNPMfTWlsDxjQjXD8SX5/fcLpmx0uqahSCIMDfG4AoWtDR3gZZ7sFHH/0agF51zkPMS4U3r5d/vsVozeT68j2jAb7ji/chbZ9MywW9+MtHDaEq/2JL0mPStknUag8XtQ6rtQaSVAFFiRi5oaJoQXt7GzyeLjgc9kW72sNXL6fdJ0nlWQXnap90FAwOGNOkONZa3Z11OOxZ189iDD05mApFEJlYf2J08rd6fugSUEZvwh+4AsG0FS6nDYAuklgOLuArwWpdVyrWr+oX92D/sHFeOTFtHKJluyFS3cfeKHiBUCZW27leijiOxX4Ppo3Dd64fmzeXQbRsX7RADIUjCPYPo6ioCB+pY9j62KMwm7ZC9gUhWrYnvb9q9K6RS8q0cfhOH4HD/iwA/W9tuYVUaeRoVCjPC5TlnoxzxBPhc8QBrIgYDYXCqJTK462YNCOHVZLKcUb2gTFm5Ek22V80BMZIvMCGh1wZY4YQHRmJoNPThc9/fhd+/etbOXN9rU89SJq+lIivLrOb9vKXx2F+dBrKXf1SJJTNQBSmYH3qwbxjRLPRFzgL+bWeJAe0o71t3vcyG+tJfIVCYSPtQxAEOBx2OJrt8zq5gmBakc83QXDWX5j+4Z0lPzX0E/0/cutXKwwxtFrdx9QpOFUHvrMq0wkAPZVAqtidsSBMidxMOsc9p4/kc2kLQlgNE4fmQPYF51xfsH8Ye/+lE0eP/ycAQMfJ/wzZF4TsC6blg87XT9XjPW843DMzMxi+/F0MvtUJ0bIdnfH2VxznkdOwfrUCYx/+ALaGKjiPnDb233LktOGKJ66z6fCryzvXk7/RCxeXySFbPRhj2LtvP6anpzLm0TmdrYYQlRbQt3Kp8OISHrIHAI+nCx5PF0TRgp64SOA5jDZbAxhj2LHzGUN4ALoQea6uEXX1jQCA//AfPADSC57yiVA2A9eX7sFXp8FXp8H7td/pbaGWKEQBPazuPXVizbXeygdWa028AO8cbly/Bu+pEwXJVSWIVNaXGH34MTC59HwWLpa4ewNg1Tl0iSSGXB32ZzH0k8jqaEaeAWtNhSH2E1Gjd43qbACQyncv+5zzSvBcIe7avurFaKLQV0ZvJt0Wd+nFSKdO/msAwN/52zF8+bvoO/cK1OhdVB34jvG8qme/M2cXicG3OtF37hXsemobgNkm+jy/mp8nvgZbQxWA2S8ZF+MC1FpTAfexN4z9hsIRQ6B6UkTtouDtnpYJL3JR1Sie+tx/k9YgXpZ74O8NwBUPn5tXyD2zWmugqlHs3bcfHk+XIXr9vQEIgmD00By+ehnt7W24NHgBfYGz8J46aRTUeE+dhCAIUJSI0Yaox9eNr9bo7w0VrGwsOtrbYLPVryvHl1j7rC8xusyLEC/I4PmLAJaUm5cPBGFrkvDMVmyyWrDWVKS1cOK/59p9ViI3DdGTC8ymrRnHX3Iht/mzf5HR0Ws6/GreJjclphE0ffPVJDEule9G37lX8Kdf/aO053V6zxufcfn1YNK+mDaOzq7zGcUp/xtJfQ/5bf7FozK+nWDaCsE0W0AlCFuN/qRMG4f72BvG59kfuLI88b+M6AjnkK0BgiBAEATU1n4FAJJmuft7A2hPCP2uVHW97vDpFfQulxOXBi9g+OpleE+dxPDVdwxXS5LK0dHeZtx2uVrg83XDe+oEXK4WjN25gevvX8PYnQ8wOHjBECPeUydpbjlBEAVn/YjRmUlgaum5T/zCmJi/COiOT7v7hVysMKdwAbFandBUEvNGOXztXNhk61yQreerMnoza0uoXPYulSp2p51npo2j7vkOozgn1dFTRm8i2D+ck4KhTCSKzxgbN0ResH8YavSucVymjRtr519UEp/LmF4IFwpHks4908bRcuQ0PN7zc7bd4iKdu9vZcmsT1wEAh+KOqcd7Hmd8QSiRm3A5bUZUwn3sDQT7hxEKRxbf7zcHk9UEwRQXfu/g3/07l74m93FDtF0avICXXU6jEl1IaY6eK0TRgr7AWVx/Xw+pCoIJklQOl6tl0QI4tSocQFyofkAuGUEQBWX9iNHpe8uaKc+FS6YQ8WoM1ac6UbnAH7iyYv0+BdNWOOzPJuUVjsQFFBfWI5Gbaa2HmDaOpm++qgu/BJeRi6W5XNVcnZs01y8cQdM3XzW+vAxf/q6+/oT1cWd2pT47qa8tFhd7/P1j2jiC/cPY8ydO4zxn++IS+kkkyUlNDJlbayoyCszE6AGgpwIAs+eK/+THPhNfFw/b83QMf+AK5Nf14qcO9wtwfUsvgOL5o3XPd6Du+Y7Fp13kIG9UksqNauyxOx+gL3B2tu9k/Od77+ljFP/pn+5AUUZzNus9n5AQJQii0KyfavplCFFgNkS/GoVnJozG4wt03pg2bjhjsi+ImDaOjrZZxzfYPwznkdOQKnanVbznitqvVhghWNGy3RA5XNikFjMBulDizljia+X9SUXLdlQd+A462l4whA6vNM+VGOWfCX/gCoL9w0mCWRC2GiI58bMz9JNImsueKwRhK2KRceM95WM+/eevGK2xQuEImg6/Cof92bSKeS4cAT29o+mwXvHuO33EKGzij/F9peKwP6uPAI073qJlO9rdLxjHOtRQBScAd8cbGPqJ7rqKlu2GI8qPV/d8BwDAe/IlvdCtfDfGPvwBLvYPGy2jzELhR8kKgimpD6S/NwCPp8sQn2+9FcRbbwUTthdw4/o1EnoEQRALYB2J0eU5IVzA7Nq1sIueMqqPQPSefMkQQcslUTDOhzkh5Mq0cYzM0TieV1t7T76EYP8wOr3nDQeKw4tJcvVaMsEFCm/xpESSZ9ZnalfE3UZrTYXx+nguIwD0nr8CJXLTmB8PzIaKc53v6/Geh1Sh51+2HDkNs7AVl97qNFy/xHSDkcjNJNGXS0TLdsMZ7jv3Cqw1FZAjQXR6z0O0bEff91/Bxf5hDP0kAm/nS0nPszVUJbVUsjVUYezDHwCYda+5cOTnO9UF5SR+mRFMW9Nuu5w2yL6g0S4tNafZWlORdOzE5yYWES6JmSmgaOX+e2MxDZJUgeZmO375y4/w9L/4PD6/W4RZECAIJpjNAglRgsgCYxrOyD6IoiVtRCmxMVk/YjRHF55UAcO0cfjPX0lzC90db0CN3k2qGF4IPAcu8cIN6GJQ9gUx9uEPFiRIH9mkv96/ecWHf77zCQBdbPA2O4kokZuG0OOi8+WUfpBq9G5uRMAcJIbqXU4bQuFImmBLDdNzwZqYt9vpPW9sx8PhF/uH4UNy3mk2EbVQmDaOM76gUdjjsD8Lb+dLxvsTY+O42D9shJkTzynTxpO6BOQS78mXdHEedyU73C8gFI7AbNpqrM9hfzbtvRRMW9F3Ln1sZ+rnLfV5S3Ul+Rc1NXoXlRW7M/YVXZGCu6ISoLgs9/tNwOVqWXajeILYqFwM9iMYHICiRNDrD2Bw8EKhl0QUmPWTM1q8uLFx8+EPXIH72Bs49MIJQyhyeGEFMOskqtG7qHu+Y87QcCgcQd3zHfB4zyflwPFCHKlid9rF2R+4krEq+3uvvQ0AeKS0BN6TL8F3+gjU6N2MbXF27dqOYP+wESI/1FCVdBy+5sT+qiuFt/Mlw9lMHRF6qKFKD7+nvFazsNXYTvYF4T9/JcmF9J0+AqaNo+rAd/BcvKjI5bQt67UkFvDwivBYgnNdGS9qch45DaaNG2FmQBdY1pqKBbvsS8GW8B4Kpq1Gv8+cN5JfJtaaCjjsz+Z3XSUrU0xEEKuFYHDA6HG7FnE02+NtyM4hNBSG++jxQi+JKDDrR4wWlemOSI4Y+kkEv/ronzES+QhAcgiZO2+JoWMuUHmrGmeGpt7z9U9MdWVlXxDOI6ehRu+iJaWA43nbV4w18Cpkl9MGf+BKWqEKd+z48Wu/mhzO52J0JV1RjmDaCltDlTF21fHC7DGl8t3oOX0kKTeUO2udXefhPvYGOr3n0eF+AX3nXoGtocqY6uM7fQQsXlXusD+LjmV2QDjjCxrpBH3ffyXtS4Lv9BG4nDbYGqow+FZnmjs++FZnmvtN5ImSLfNvQxBrlGBwAC3OVgyFwgUdWLBcBMEEm60eNlsDZNmHYHCg0EsiCsj6CtOXfmbJM+lT8Z0+An/gCvrf+QcAyeFj/nti6DhReJ6J58nt2rV9No8x3p5IqthttANKJbF4h2nj6PSeh62hymgQLvuCRrrAv/+bv8Z3X3s7SSRbayog+4Jp+aM8L9NwQAs8+73v3CvG+MhUxyxV1PFm/h7veT38/MKzxjlIDDnzsPRi8m7ngr/HHe4XjCKhRNdbtGzPOIudWAWUUK4msfrh3RdCQ2EIggmVUsW8U6MY0+A+egyHbPXrZjxnj68bVdURtDhbMXbnRqGXQxSI9eOMAsAj25b8VB7S5YKDaeNwd7xhCJtEMapG70Kq2G2IutSJN1yY3kpolG6012HjhjuYSqIzeiY+qjGxHVLiMbjATJxqxHMIM6UKcOGUWDBkPC/+OlqOnM7bSFHuKi4E3+kjuP9f38aNf/DNKwBzlYOY2s4oW+N7YpVR+hkSo6uM4g/DKLq9dh28lSAYHMBzdY1osh+GLPvg8XShyX4YVdUH53ye3x+AqkaNyV+Fgk0UwTloQt2bT0DVlheRFAQT2tvbwBijcP0GZn2J0RITULq0CxFvOeM/fwXK6E24O/SJMK5v2SCYtiaFjlMFnf/8rIBLbP2T6lry/Yi79GpoZfQmlNGbGdsz9cbzIrmwlSp2p4nMVHeVC7FMVeS2hqqs4WvRsh19516BWdgK55HT2PzZv8jb5KDFkM/pUkaFf/y9SZ14RaxCikqAss8VehVEAqV/dxybvteIslcPoHhk6WHY9TYhyt8bgChaMHz1cryH7Tk4mu1QlAiqqg9mfb3BYD8kqaLg8+Q7f/oY/O9tRuj2pmWLUUDPIRVFC2TZt6ZTD4ils77EKACUPb2kp3G3kmnjqHr2O/AHdDH4stNmFKtwMWIWthqOZGqjeP/5K4ZYTXXXePGOEtFngPN/vNcih1e3q7fuwh+4Am+nXj2tRG4mCVK+3qQ55JbtWRvB8zzHTNgaqnD9Zz5c/5kvY0X+RoMXH/Fzu9YmXm1INu0EilfnONxcMp8wWy0X80f8rSh914epajseOrpRovQvaT+y3IMdO5/Bjp17EAqF53/CGoDFGCqlcgiCycid9Pm64XI5oSgRPFfXmPYcRRlFaCic1O82bb9MM4qbVupchW5vgvyLLbA9PQHXl5Y/7Yzje11PO2hxtuZsn8TaYf2J0eIy4NGlCdK+c68Yjbvb3S/g0ludeiPulDnc3pO6qNz82b+Au+MNiJbt8J0+AgCGm2prqErrm+mwP4vhK9+F7/QReE++ZPzjuadcAJmFWeHDi5M63C9AtGxPKoJy2J81CnyM1/D9V4zq76XA+0uu1hn3+YK3ZUptNUWsUjbtADY9VehVrCj+3gCqqg9iz979Wd0zHv7du29/Qd3E4pEBlFwNYPqpCjz8q5OYeXJxo0sTkaRy9AXOwWw2oa6+cdWI7eVQa62BvzeQFpb2njoBm60BihJJq5bn4tJam1mMBoMD2LN3P5rsh+F0tqKuvhE7du7J+fny/FS/Nri+fA/S9kkoH6eXnoRub4Jz0ISmi2b439u8oP1arTUQRQsUJbImJ5kRy2P9FDAl8sg2YHoceDC26Kd2tL0AtCXfZ4yDvHUXUvlu2BqqjCIcQdiKDvcLkMp3zxa9tL2AYP9wWjU9oFeMZ2pzk9ijUjBtxdiHP9BzU+PbCqatuP4zX1qoPrVv5Gpr7bNW4fm3/vNX9LZNc4wdJQrMph1A2ecLvYoVxd8bgNt9HGazCZVSOUJDYXR6uuA9dSJpO1n2wWqtQV/gbIFWqlP6oy4AwENHN7B5eTm83AmUpHLs3aeLrevvX1v2GjlctOdzSMHLLic8nq6MwtLRbEcw2I+hUDipIbyi6FGabCF6UdQF//DVy4jFGEJDYXg8XXiurjFnhUHKx6UI3d4E29MTsD71AGyiCGd+nlB4O1GElkEBodubIJqmoGolCP6qDJXbHkLaNv9gGl5Z7/F0rZsCLWJhrE8xCugXp6JSYOL2snfFhWJijqatoSqtACexuMbWULUoAZPqRPLRiKls9PB5vkicRuQPXDFSLmIsN9X6RA4oKtH/zpdRuLgWYEyD230comjBpXhz8B07n0Ew2J8kRnkYt729Lduu8kLxh2EU345gqrIBM0/pwmn6qQqgenn7FUUL2tvb4PF0QZZ7Fjx0gDENqhqd/XcrCmUkghjTwBhDLKb/BHQhl498TEEwQRQtiMWPmwgXlbEUZ3sk7nDyx1ORpHL0+LohihZIUjms1howphltk7JV6vPzM5LgSB6yNWQ8D6Hbej9v2x9O6K+jbAYdX/m98fiZn2+FaJrCjZc+hlA2A1Urwd43/gCdP30MfYdic54TALD9q3rIsg8XgwPwzbs1sZ5Yv2IU0MN2RWXAp79a1m5SRx0ulEK3UCKWR7v7BfgDV4wxoMBsCgVRQIrLdDe0dNuKjvxcLZyRfWCMoS/wDgTBBMY0CEJ6Y//5wrj5ovgDfR1TVQljHjebMP2F5a/rZZcTvb0BuI8eg81Wn1WYJVJVfWDBYV9VjeatOEjcZckYQr8Y1HNrU1+bvra5rympgtP17RbIsg+hoXBGMcqYhqamwwgNJeeX9vYGMrrPoaguRq1PPci6ho6v/B5C2Yz+GkxT+vNuL2wojdVaA0EQwBiDqkYX9P4S64P1/z/5I9v0i9envwKmJwq9GmINwWe5B/uHjXZe5IoWAD7es8Sk/z0XlW0IEcoJBgcgihbjwuz3B8AYSytk0ftVClkLXBRlFBeD/bilRiGYBbi+3ZL7i/19DSXDAcxsFjD9zOLFZ9EnUZRcDWDyWWfG8L4gmNDcbIfH04VgcGBB7qj31EmMKBF88MEv8eabb+Pf/Jt/jaf/xW784G/78GcHv46DB78Gs1kX9/muUg/FG9dLUjkY09Dp6YLfH4AgCGkz23X3dmnvF4ulO7CAfj4TR3HOl7KgfFwKoWzGEJmpJLqkxjHiwnShHLLVw98bWPD7S6wPNsb/6CUmYOuXgClNb4qfo8b4afAJUMVlQNGjWTaK583MTAIzU7M/iVVJz+kj2PMnek+/fEyoWlMsZ+JZopgsKonfLo3/HhefxVvjtzeW+MyEqkbRGc8x7PR0QRB0MZmIokSyiktVjaLJfjjptiz7MDhwYc7q7MVS/EEYRZ9EMVXZsKRc0Uf8rSj+MIyZLQKmvp5ZiFTG3cFOT9eCxIo+5aceweAA3nzzbfxN6/+stxF6zYdNZZty+voXg1RZgdBQGFXVByBJFUZOqCAI6PF1Jwlj7uyaF5jXypiGkZEIBgbfAQA8unnuIiIuQmMxBsY046fuVOrHVLUSqFpJmivKJoqgaiUZc0JVrQRsomhB+aIcwUzjfDciG+t/+BKT/u/Rp3WXdPqeXugEZBeFqRfc4rL4L6X6BZJfPI0L6TJOKV/D9L2E2xMJwnUieTsSsisOLyZTRm+u3eKwxC9JwOwXpeJNCfdn+DwnPhfYsIKQX6h5ziGQHg5dSazWGihKRM+VjIcwvadOJokoPs3H0ZxZWDGmQZIqjMKmUCiMuvpG1NU35jRPsug3+vmZlhZ/fopHBlD8oR4uLvoke1idv27GWDxlYWECjT8v0XG7tcSq7bq6Rqi39M+DIAgwm03wvd69KGEr7pr94qAoEcPV9p46kfalwnAs5xBqjGk4E88P5fszm/VzU/OV/3bO5+3Zu9/ImwV0QSyKFnSgzfism8umk56naiVoumiGqpWgcttD9B2KpbmgPDxfue1h1uOnwt/PTPm0xPplY15dgLjzUgbgiUKvZJaiuCAwBO8iMIQpF6pTCffHRez0xOw20xOUtrAIci5EU7/kpDqFme5PFYjGz9KE58Q/P7n6grSBSJx+wy/MicUtiTia7Xmr9k0VWz5fd1oId8SotM6cU6i3R5qtsLdaa+DzdcPpbEWnp2ve6vtEocPPhyRVoMfXnbS+4tv6OmY2L9Lduq8ZFfjzbOvQpQAAIABJREFUIQgmw0lcTI6n3tNTQGgoDJerBYIgpBUJLRSbrSF532bTolMe+PaOZjtcLuecryMWD7PP54wOhfQ+pD2+bjzxhIA/e+55CIKAhobsk50EwYQb168hFApDFC1GcVU2Rj5+RF9L2TSsTz1IyhFNhE0UGW2gmss/nXPdiSjzFGoR6xO6Sq0XuJDFEoQsd4m5cJ2emP1JgjUzqSkZieHlbK658Vz6s1uN8JY63PFKxWQy4fOf3wUgXYysJDws3d7ehpddzoxO4FIu4IdsDXCidUHN0WMxhltq1HD+xF0WQ7gkcT8u7rYsLkRfesWH4tsRzDxpmdMV5fB+lLEsuZDZqJTKjddrFkxJr51Xlc8nxgDA5WoxCmwY0xAKheHvDaBjEZ0M+LkcieeMzsVCesam5n8GgwNQ1Sja29vmdY954/05t4nniqpaCZSPSyFtm4T3a7/Lur3/vc1QtRKjDdRCYEwz0hUKXYhH5Be6KhIJLvEcTI8D0w9SBGvcZQXit9dA2kCim5h427gvfn9SCBvJgpNyGNclLldLUg7ijp170Pa/uNDS8k00/qUDpSUlSRf7hcKLVGKMoVKqWHSIn4sWRRnNKiq4eJ5PRDGm6QUy8fDtk09+Bp988hv4fOewc+cOSFJ5xn2IogVe70mMjESM2+YMIeOie7o4nPmMvo/UufRF9xgm/zxZsBV9EkXJuz2YedKCqSr7ghzSbAU588HzNHlR2EhCNXtqVbnL5Uzr48rpjLeXSmQxQhSYdWoX0+B9MU6uLPuSCqEUZRT+3kDW17QQmr/4KTxXt8L948eNsDybKAKQXKikfFyKzp8+BqFsBq4vL3xK0xnZZ3QNIGd0Y0FXVGJhFG9d/qjFmYUnsS8aEofECvD/Xvs52tpeRmlJCUJD4UXlKALxHqFHjxtuDzBboLJQUWqEl+dwMGeLX+ZfWzA4O5ZzfFzPmX/n8v8NQA/nzyWqPB5dKPIenX2Bc0muHs8ZndmiC9WSd3tQMtKPmc0CZp606JOY7mtJxU2l/4c+heiho9toCzVf8ZN6a2HiOxVrbQ1k2QdZ9kGSKpLyTmutNWhvb4OijEJ+Td+GxVjGdAxrbQ0UZdTo6XnI1rCkpvnc4Z3vc8UfW6gIDwYH4ukIztm+pTEGWfbB0Wxfco5wx1d+j+CvyhC6vQk7/uN2OL54H4DugvYdisH29ITR+J5NFMH7td/BXDaNzp8+hltaMXx12cW0oowaAj81DYVY/9AVnMgfJBiJNYQoWoyLvyhagCFgZCSyqCIVQTBh+Kpe0ayqUUMktDhbceP6tQULGD51KVvvxVhMv8hncitT18PdXUUZNSq5FzKxKTUMnImZzQKKEEXRPYaZzSY8dJ7FQ8THgw4HMPmXJ5KEJherk99ow/QXalA8MqDvZ0v21xEKhZfsnvHepKGhMKRKPf2Bv6dchFutNXC5WrB33/6szdet1pqcVOHbbPVQ1ShiMTbnZ6GSrzVeqDbf6+auqOvbLUYagfvoMQDL76U63PwJnIMm+N/bnDTqM/jLMtienoD7x48bI0L9o5vh/vHjAABp2yTYRFHm/FKmwe0+prcsq62Bw0FidKOx/mbTEwRB5ACzYDLCuLyKeSlzvhnT4HS2Qn6tB4LZBEezHWN3bizKSauNC5/U5uTLgTdX55N6/L0B1NU1LmuffOISr4rnlL7rQ8lIP4qVAeO+R/ytKP1hFya/7kwL3c/ljHJ3dqkdDQYHLsB76iQ62tuyOnCJOZoLyddcKh3tbbg0eGFeccnzWBljaTPrU2FMi7v4DFXVB7Fj5zNosh+GqkZhrc2NiPbVabj/nf+Kwb/+LQb/+rfw1Wlo/4rusvMCJ2C2L2l79Tgu/fVvsvYc9fsDRp9cr/dkXkezEqsDsqoIgiAywB20bDl9vOBlIS4T348gCKiMj2pczAWXFzHNJ4YXuh7GNPTGRU0oFMYev97aZ7nh0ekv1KDkagAlVwOYqtb3VXR71BCnxR+Gjf6hU9V2TFXbFzWZyd+rixZRtOBll3NJaxRFi5EbnK0jQqenC6oahW2J4ffFsFCX0nvqJJrsh+HxdM2ZeywIJrhcTj3NgDHjM+dyORf9uZuPTIVJfYdiCP5Kz7WXtk2ictvDORvfy3KP4dp6vSfyPniAWB2QM0oQBJEB7obyCmsASbmfIyMRPLcAJ1EQTLj+/jWM3fkAY3duYHDwwqIFgc1WP2f7n8S8wIXAxRYw+5q8p07C6z25qHWlMlVtx8yTFhR/GEbJu3r+Hy9IShWd01+oySpEi385jEd8L85W50MX4m73cd09O7Vy7pn76HHIsg+iaEFPntp3LQS9eb/exaHJfhhOZ2tWl9R76gTu3xtL+szZbPV5cRxF0xRcX7oH15fuwfrUgzmFqL83YAxxyNSyjNg4kBglCILIABd+7qPHDWcysZrZbNYb0C+kNRKwsOKiufCeOpH1Yl2ZsFZZ7jHyKrMRCsVHh9bWoC9wDjeuX4v33ly+WHno0AXcI393DI9+ewdKRvoxVW3Hg7+5gIfODLmp9zUUjwzoYfv/5/sAgJJ/eNuoyE9cc6VUvqjir8VgpFPE8y19r3evunBxX+AsXHFH2N8bgNPZiib7i1m3X23rT8XtPm6IfhKiG5ui+/fGFjc4liAIYgOgqlHs3bcfklSBS4MXsGPnM5CkCqMgiTENO3Y+g/b2tkW39ck1fC2JCIKAsTs35nzOSokV7owW345g5jMWPPjWuax5oJu+16iPAN0sYPrp/Sj+8Coe/NsfYOYPq1Zkbdlosr+IYLDf6HYAwCgWyufErYXCvwSJomXNtkEKhcKorKxY9aKZWHlIjBIEQWSB9/YURQtCoTDMZiEpVL53337EYtqcoi9f8B6igO7qpq51NVP0SVRv+1QgeGeBTORz2hZBbFRIjBIEQSyRYHAATfbDGLvzAbk7a5zEHExe9NPssFP4mCDyAIlRgiCIZRCKzwMnCIIglgaJUYIgCIIgCKJgUDU9QRAEQRAEUTBIjBIEQRAEQRAFg8QoQRAEQRAEUTBIjBIEQRAEQRAFg8QoQRAEQRAEUTBIjBIEQRAEQRAFg8QoQRAEQRAEUTBIjBIEQRAEQRAFg8QoQRAEQRAEUTBIjBIEQRDEChAKhdHp6Sr0Mghi1UPjQAmCIAgix6hqFFXVBwEAVmsN+gJnC7wigli9kDNKEARBEDkmGBwAYwyXBi9AUSJQ1Wihl7QkFGU0r8dT1Sjq6hqxecsO7N23H6FQOK/HJwoDiVGCIAhiVcGYhmBwoNDLWBaKEgEASFI52tvb4Fkj4fpOTxdkuce43WQ/DMa0vB3fffQ4RpRRtLe3wfVtJ5rsL+bt2EThIDFKEARBrCr8/gBanK1rWpDGEgTcIVsDQkNrw+FTlFFjraoazauj6+8NIBjsh8vVgo72NrhcLRBFS17FMFEYSIwSBEEQqwr1VhSHbPVocbYWeilLhsUYBEEAAAiCCZJUsWZC9SzGCnLcYHAAgiDgZZfTuK9SKi/Iect3esJGh8QoQRAEseqIMQ3DV98p9DJyhlQgUbVYzIIJ6q3kdeZj3XpqRj+s1hoIgsm43+s9CUkqX/HjpyLLPtTVNeb9uBuV0kIvgCAIYjXCmJYUplTVKASzCdbaGgCA/FoPvKdOFHKJ6xoWYxBFS6GXsWRiKaFlR7N9TYSbBbOAWEz/7JvNurMby4NTejHYDwBpwjNRmOYTl8sJf28A7qPH6e88D5AYJQiCgB6Wk2UfRpRRqGoUjGW/AP/Zn30dE59O5HF1G4+RdRAmTfwMrSVhzRjDX/3V/4QLF/wA9L8Nq7Um47ahUBiMabDZ6pd1TP6lr1KqWNZ+FgtjGs7IPijKKCSpHB3tbQB0Uew9dWJNuNnrARKjBEEQ0IsnYkyDKFpgFkyYnJrCtWs/x59aa7Bz5w4IcZcoFArj7//+XezZ84cFXnF+YExDKBSGJJXnTVAlumHcoSuUQ7ZUzGtsvYD+Xl+6dAUA8O//1+/AYnnKyHvNRCgURl29Hsp2NNvh83Uv+dg8R5M7o/y2qkYhCCaYzUJOwvWy3INOTxcuDV6AJJXD7T6Gi8EBuFwtsNbWgDENgmDKKrJVNQp/bwCMaQtyTPn+iLkhMUoQBAGkXVj4hdbjOZ52ETz84r9FODycz+UVBMa0eFV7PwRBwNidG3k5rl5BzRCN3kaT/UV4T51IcuYY0xBboTC+M1401d7elpP9rwUxwpgGvz+ATk8Xtm7dAgDYvu0PjMd5myp/bwDiLovxXng8XRAEAWazadlONi+aMpuF+MCAA2nbSFIF+gJnYTYLaHG2QlWj6AucTXqf9CIoE6zWGsP1dDTbIYoWKMoo3EePAQBGlAgkqRz+3gAczXbDEeV4PF2w2RqS7lP/f/beP7yN+77zfIM/RcbGsFbc2E9FI1nHP8kDkt65YnBHnFWpJpkE6h6jEM/uErfSZgnfneHSe0vwNqWo7iMSm67Ie9asRrs10K10B95uwai6RlOboFeOvMMWoaptHE5JW7bia0Zg1koVOfjCiRhS/HF/DL5fzuAHfwEgQfL7eh4/MvFjZjAYzLzn8+P9UWNoam5l0VJRDOD2Rx9k/X4VZRovNLXiqLM5J6G+F+ANTBwOh5MBWt+XqV6uorwcicTPt3qTcqKpqXVVqyRJiuCRR58yvCYU0qx2AKCmpnCCipAERDEIX/cphIbDqElG4073DaAmKSz0rz0rBuDrPrXqMmU5yj6LokwbPtdqtZvt7S6EhsNMlO52VDWGRx59Ev3+QZzs6YK/XxNrtInJYqlFnCQ0Iec7xfY7IQmDBVSuwp3W2AqCGRZLLWbv3caN967j2sRbGItcwkj4ArzeDpSUmNgNks1ahzbXCUMqvd8/COnPte9alqPw+wcRGg6zvwFg4EwfjiaFpsVSy2pD9ctRb8UQTynV8fsHoaoxuNtdcLe7tO3WnR/6/YPsOFPVWNKjlSA0HObp/jXgYpRT1OSj4L/NdWJH+xVytpfUzuJcUJTpbWtisdrq0eHpZOvXPB11wnM4DEIIixwBgHguwP6/PXnxLQQ+Xy983b1QJqcwHArjG//0ZQDAxYvfQbvbuN6DDYchikEmBjIhikE0NbeizXUcoeEw2lzHmXl7aDiMRx59Muv3YLNpNYv58gVNvZkRxSAeefSpjOvfDsFisdTixnvX8f6N6/B6O1hklMRXtk+Wo/D5ekEIYdtIPxeti/W+1JHTdtCyhtT90uY6juHhMJzOZrjbXfjBD6aSXqQeOJ0tUJQpJjZVNQZFmVppMhQDhmXSmxr6egAYCV+AxVILUQzg6WeeY2Lb8litwd5JUaYRGg5DEAT09HQx8UyXLYpB+P2DEMUACEnA4+lEPJ6ANVkDu9bN016Hi1FO0SHLUTQ1teJgwxE88uiTqKp+BG2uE5saC0ftQjicjUJT8/qLsh5CyLqFJb04HWw4jEcefdJwMdwqTvZ0gRCCs8kLtN8/iDbXcSYuaCqWIstR9pyj0W7wfswnkhRBaDgMr9eDsbFLGBu7hL/+L1cBAPv2VbIIFoU2l2VrqJGkCHzdvSx6RaNZVms9BMGM4ZAmKGhqNfVmlZ5nrDk20tAaY/0xQkgC/f5B2Kx1htSuLEfh6z6Fgw1HUFX9CNvvshw1TEMiyQglRWu6C+Z8g2Ox1LLtof/SqGCNYAYhBPJ4FIIggBCS5sHp9Xqyfh/rxZq8CZic1I5DQhIs6nlZ9/1QgelORrABY30p/QyiGDREbunn9Ho9UJQpPPX0c6xp6cZ71xEIDMFqrYcoBtjxaIiU6pYhnguy64qqxiDLUfQnJ2zJ41EcbDgMeTwKr7cDb45dAgBI0ig8nk4+3jQLXIxyigZFmcbBhiNoam5l9UdOZwscjXZI0iiamls3fBHnqRHOZmG2Nqt01VNEMYiq6keyGmX3J1OFTmcLBEHAePKCpCQ797cCQTDD3e6CKAYNllV+/yDbDkEQMBK+AGAlImy11mNk5ELB6h714oLyB2e1x5544vGM66XCMhUq9iyWWgwM9MHpbEE8eTPhdDZDkiKQx6Mspayl70fR7x9kaX1qtL9a5HWzTE5OgRDChBdA07knIIoBVgrB0srjKyIHADo8nXihqZU1lb3Q1Apfdy/a2o7nbRvpvrmVPD7otgqCgGCy7vFgw2EcbDjC3hMKhVmJxWbFlvOrWqNQU3Mrmppak2JxiglgKrgnlWlYrfXJ70sThHSd0e/9lfZ3cr/R5itZjiI0HMZjljp8+9t/xpZJv3dRDGJcd/OlqjE4nc0YCZ9P205FmYIoBtjNihZ110aWepM3bPF4AgNn+nCypwuCYMZY5BIEQUBoOIym5lZuqJ8B3sDEKQpooTeg1fO43S7DxUYTqofh850qyEWCw0lFEMwQBCHjhYNGvVQ1hnicMMFwWRpNa3bSLnYBeL0euNtdOCgdxmPJEYcdnk7YrHWsuYF26tKGi3xDvRP1c9JDw2H2eRwOe1pE2N3uKmgDjjwehdVaD6u1jolJUQzgmWeexAcffJjxPdn2jSxHoShT6EmKgNTvgn5PVHSsRNamWFc4oO0nrze3tDPdxv5kZFZv86Tfn7Q8YuBMHyyWWrS5jMKSvo+OyqTLDIW0bXc02iGPa0I6V3slQH8Tpn3/VFAddbbAaq3DWOQSO34aHXaMy1HI41F2UyEIAutU3wgOhx0DZ/ogngtAHo/C0WiH1+thy9ai23XJyCyBr3sKVms9bMkmpDbXCfzn//yXAFYaq4KBIahqDL7uXng8nfjCF6wwP/gA1Fsx2Kx1cLe7MDwcZseBo9GOdvdKPaj+OHM6m9HT04VbySi719uhCfBQGDZrHXp6upKf4XRa05rDYcftj96HokyzmliOES5G80zp1SBKJ0exXC3gvnsIqCruLspiITQcxlFnMwYG+rJEQurgdLZAkkYhy9F1p4S2wqyZs3tZ66IRjxP4/YNMMGQSrsPJC533pQ5Wc/ay14NQKKzVt+mOZfFcEKIYwLgcxVgyvZdPrNY6WK31houvXkjo6/7oZxfPBeB0NhfkAkqjXaoagygGERrW9onVWo9v/JN2dL7yL9DvH8QtNaZlSZL7Ktu2UJFJDeZ93afYd6OlswmLiml1m4SJnslkmQIVXbkgy1EWrZOkUTga7Uw0AWANWgCSqWJN3NCmKVrzSJGkCHy+U2zbaWRuJHwehCRwsOEw+v2DeRGjVCzpU976TnOHww6HQ3ds9tDSiQSzYNrszYvX25F2E5BauxsIDMHnO8Vu4mpqhKQYH2XfpSSNotFhN+yPOCF42etJ2zZnstN9Pa4HqR33A2dOZ7R3yrac7ZgktVPgYjSPlIc6UToRxtKBepR8HEP5xV5NkK5ByWQEy/trsXxg7x6o7nYXXmhqNdhftLlOwPtSR5rwpHfuG2E9qVYOJxWHw56x5phebPz+QRbFkcejGWd6x0mCzSUXxSBO9nQZoqn0PbIcZdEueTzK6tnyTTAwxGxzrLZ69PR0wfNip0HsAdpF2mqth6JM4elnnoOj0Y5AYCivopTObFeUKdY45W53YWCgDwDw7/94GH7/INztLsO2WR7LvA207pVa/9BGF1pD6PV6MHDmNHzdp1g3Nv1MuQo52uUvSRFD/e3AmT4msIQaAaIYMJyPFGUK8XiCpbmt1nr2WR2NdvgBFi0dONMHocYMEk8Yske0fnIjN+qrEXhtiN1ErYdCRvq8L3UYuvX1neyUG+9dhyxHYbNp5Rup3+V6otzFbr+12+FiNE+UTEY0IfqEHfMvXoDpriZG1/O+isBxLNpacN+TXp+yV7Ba6/D+jeuGxxRlSrvQO+wQRa1g3GKpTXY+Btd1AaGRl2xNKBzOalitdUwg6qGRLRq16enpwqTrREavRdrh63mxEyd7uuB2u5Jd7VqULk4SICTBUp9erweiGECb6zhuvHc9bXn5+EzeZGTW+dVmOBz2rOu5NnEFohhEnBDUCMKmbgTXgq5DvaVZ5ugF+LWJK1CUaUODzUj4QlbB9VhSsBxsOAxVjcHr9bA0sz6qlS2itVlouhYAbNY6jIQvGCLOlEw3F45GO0LDYYhiAIIgGLbL4bCzRhraTZ4J2nyTL7ToZ+6iNh9YLLUZazdTKZbt5WwOLkbzwWwCZW8MYrlqJTW/fKAOC4fW7j4tu5o8WfF0ftqdaTyewNWr40kPQu1EHXhNi5z6unuT1h6ri9Fsd7s7wYias/1YHqtlzRP648VYS6ZFFI86m5k/5WOWWtis9XA6m5lljaPRzia+0IgdoDV/dHg6IY9H0dPThZM9Xaz2samplb0unxOQNiLGcq2dzHUdqQJutd/8yWRN36QyjYEzudd9rhdJGoXb7WLCF8icjanJMM2ICsk4IRlrhYvpe+JwCgUXo3mg5IMoSmamsPDlLizvXzmRLNlWF0olkxGU3NQiK8tV+Y847GT6/YOYn59DNHoN0eg1VoyuT1+lnqRXE5j6CwNtHLFYavMaHeHsPqjnZDxODMeW09nMIla0w3hgoA+XkzZFADAW0erqRkYu4IUmzQlCX6tJI3WiGGDilD42Ej6Pgw1HII9HWfTV0WgvSB3pbmM7Jt2MhC/gYMNhFmkGVoSn/txDI52pdlVcSHL2OlyM5oGSGa0+aLFhA13eswmU/+naafy9CiEJVFdXY2lpGWfPnsFRZ4tBDKRelKnAVJQpQ8oxUyRJm5IxysQCh5MNQTDj2sRbGY+j1BsZ7bVXIJ4LGkYm0sfpFBZbsiaQHs+z926nLdtiqWXvAbSmGt6BW7zQLvOm5lZcm3jLEM3Ve4AKgpnfAHM4GeBiNA+UXgtrDUj713+xKPtuAKa7MSzaWlA6OcrT9CnQE7YoBuBotKdFPKn9DeVsMrpEPRIptMaNdjlLUgSiGMDjj3+O1xhx1sVGmohWi7Zv1JLMYqlN697lFC8Oh9bg9UJTK94cu8TOWZma2vY6vEyKkwo3vc8DprsxLD+UFKKziZX/slByM4rSq0EsPWHH4kHtArURIbvXSDUFV9UYPLqxhoAmMgVBSKsno16RshxlHoYA8Fu/dajwG87hcPYU7nYXjjqb8UJTKxsaEM9xOtJuQ5IieOrp5/hAEo4BLkZzJSk6qZgsv9iLyt7nUNn7HPa98llUdj2J0qtBw+vLLp4Cqs247x6CaZbfNWeDRqRSvRtTT2LaiLwpHNUJUUmKrHgOul0ghLA0/v79D2F09D/xKRgcDifvBAJDsFnrmGcox4g2u50YRrByOFyM5ojpnlFM3ncPYW7wfa2r/v4vsXjIg8VDK8Xp5aFOmO7GcP9rfYZoKG9gSidTA4AeKjbpLGNq6yKKQXR4OlmK/mRPFyyWWkjSKDOKBjT7F37B4HA4+WZk5AKr8a3ZQelomj0qZNRSvaV5hqaa2W+GkptRlL0+iLLXB9d+sf59kxFUdj0F010enS0WeM1ojixXJ0WkPi2ftHpKexzA8q/VYf7LXZkN7mcTvHZUR7bIqMNhN0wo0c+Zpn5/Xm8Hq0kSBDNuvHfdYCJ+473rkKQIN8PncDh5RxDMGAlfQL9/EO07aHxxKBSGJEVgs9YXrGFOTY7TzEXwmmamUR7qRMnMlCGQs/CVddRYJ6/PplkC08w0L5ErErgYzZUqM5b316Lkg5W7vLLvBliHfWoaXv9jKbkZRel/0Tq6Ky78b8D8PfbcvOfCmtZQu53VTob6bnqLpRZWaz2blOP1evCyN93jdSN+hRwOh5MLVmvduszaiwlFmYLNWlewc6NegNYI5k01MplmplEROA7cS+D+sT4sNri0EdzrzC6WvT7Irs97eephscHFaB5Y+HIXykOdqPjWESwfqFsZCTozlfU9dPIS9j0IAFje/xiWLDYsP6R15S89yTu9Ac2rcT1TX65NXIEkRZLClJ9gOBwOZ6PESWLLygqstvpklHQD5+vZBMov9mqlbu6hFTvF2QSWbC2rvxfadZcNmgFvHC4muBjNA4sNLphmprWDfGYKS0/Ycd89hJKb0ax3a0u2Zsx98y3tx8BT81nZiP0Sj3RyOMVB6ghPzvYgy1E2Qnk9KMoUnM61Rd1mSfVcleXohsRo2XcDKLkZxaKtxeDrvdjgWvM6arobQ3lopUdg6UD+xqdycoc3MOWJhWOn8ctztzF3+jrmX7mE5f21WGxwrZpqXz5Qx4Uoh8PZdfT7B9msds72IY9HN9SkWWi7pbjOczXTaNTVMN2NofRqEMtVAha+luLlu9Z1dDaB8sAJAFomc13v4WwpPDKaZ3jYn8Ph7HVkObphk39O/jnZ0wX0rO+1VIhuJppNSAKTk1MZM1mEJJJDSSIguobRjTaPligRmGYJFhtcG77OMhcb9xBQbQbe0DUfc4oCHhnlcDicPYaiTOPpZ55j40bzvWxCCE/R7zBoCn2jEcvQcBgHGw6jqbmVDRXRL7PD0wm/fxA1ghlWa33ycYJbG4zCltzUmoSXnthYP0Xp1SBKPojivnsIS7bmFTsnHhktKnhktFDMJlA6EYbpbkxrZJpNGD3Nqs1as1K1gOVfq8PSgXqtaamIfyD0DveWGkNP0ruTw+HsPOiY3UJEL+N8/OW2oCjTEATzus7LijINn6/X4EpCvzehxngNEsUgxHMBXJt4y3CDQUgCPl8vu6GxWuvTjqcOTyckaRRerwcDZ05DUaYhSaOIxxOQx6PwWtNdT7JBPb31UVHTzDTKL/Zq19ekc83SgXos6OwTS25GsXDs9J53pyl2uBjNI6a7MZROhFGiRFasI6oENrfe8CO6R2D6OKbd7U2OsseXnrBjscFlKM4uBhRlGh2eTqhqDDU1ZsA/iEBgaFu3SZIisFrruCjmcDaIPB4taKMKJzv5nstOo4/U2m4kfIE1c0pSJK2xk4pIIcWl5C//ckJ7Pr7SZKQo0yzaGY8bo936dbrbXRgY6DM8T4Wn09mCgTNajSc9VxNCQAjZWNMpnXagBiNJAAAgAElEQVSobwquMmvp9v21WKLX2gN1aWn8jUZTOVsPF6O5MptAycwUyi6eguluDKZZwu7MFm0t6+qWN81Mo2RmSvMdnQhrUyUunsLCsdNFIUolKYKOZBF8MDAEp7MZspz79IxcEcUA5PEoxiKXNtR1z+HsdTZsqbMBCJ/FDkISCIXCcDjshv0sikH0+wdhsdQiGBjKy3dARaGj0Q55PApRDDCR1+HpxMmeLni9K1MAz+rOmxRFmcb5C/9P2mfo8HRqo5Q7jhtu+kUxyIRmjWBGo8OeJrDpNUIvOAXBDHe7C6HhMKwbNdZPXkdLZqawmIx6Lu+vxX3PeZTcjKLkgyiWD9SlCc/7nixer7P8OC0muBjdLLMJlHwQRdkbg2wKxOKhDk2AbtBId/lAHRYP1GGxwYX77iGUXg2i7GoA5aFOlF4N4r57aNvMefUnz5HweXbyKAbx19PTBfXFTrS5TuD2R+9v9+ZwODsC2qiy0dpAzgprRTfPigH4kxHFQGAI7nYXRDEIX3cv3O0uXJYiuCyN5ixGafSRrudgwxGot1bKwSyWWkOjkKJMQxSDcDpb2DlcUabR5jqOp59+ErHYj9lrOzydUJQplJaWou/07xrWK57TvDpP9nTBaq3LeANC16ufoEdIApPJv8kGG5gWbS0sULNcJWD5QJ2WXZyMaCVxswTLVwXMDa5+LaCR1ZL3/1IrnUuW0JlmifbvxzGtY//Y6VWXw8kvvIFpE5TcjKLi1VbNtB7A/WN9mOu7joWvZBnzuUEWD3Vg7vR1LHy5CyUzU6j81mGUTEZyXu5GoSdPi6UWb45dKrp0uMNhR+C1IRBCIIrB7d4cDmdbUZTpdVnz5NI1vR7ocvWiaLfxQpOxWUdVY4Z9P56MCvb0aDZCdOa7o9GO9nYXCCGwWXP3udTX57a5ToAQkjYqWZJWrh2XpVEQQlhtpyxH8UJTKwRBwD/6h18HAEjSKHzdpyBJo9i3bx+qq6sM66HHmX7AiCCYoaoxSFIE/f5BiGIQR5NlIKIYQFX1I3j6mefw1NPPQVGm4Gi0Q1VjG2qgWzzUgaUD9TDNElQEjqPy1HOoeLWVmdgv769lIz4zUToR1l7/xiBQUgrTLxOoPPUcKr91GBWB4ygPdaLsjUGUTnBLsu2Ai9ENYLobQ3ngBCpebdVsIo71Yf6VS1g81FGQxqOFr3RpxvhVAioCx1k34VZAhajVWo83xy4VbWesw2GHxVKb1sXJ4ew1+v2DaHOdWPN1VCSuZ7LZRpGkCHp7/QCA6F9ey6mch5AEmppaUVX9CJqaWg0Rtu2mRjCzG2BCEmhzncDBhiMANLEmj0dhtdbjZE8X3O0uyHIUhBB4vR4MJ1PU+RjSof8OFUXrU2jXNRF5X+rIKJLjhKCpqRVNza2oqTFjJHwezc1HYLHUslQ/APz+7/9LfO5zn00KXS36Sev0VTWGRx59CgcbjuCRR5/C0888hzbXcYzLUQg1Zlitdbg28RYcjXYIgsC2Y+BMH8bGLjGhvhHmv3kFC1/uwtITdm1SYbIkbq7vOhYPap8723WyPNSJkptRmO7GsFz5KWDfA6yXY+lAvbasQx7MffMtHhXdBniafp2UTEbY9IbFBhfuH+vbks735QN1mP/mFVR86wgqXm3VxGmBU/Y7RYhSHI12hIbD7G6dw9mL0Fq8taANKqm/ayo2sv3eZTkKeTwKQhKsIUX/Xlq7aP/SbwAAxse/h8XFxU2X9IRCYcjjUYyELwDQxLZ+1jshCchyNKuoS02ly3IUNlt9Xs5n7W4XZE8nQsNhWB6rZUJQv//1n1seXxFIoeGwoV4zn3i9HsN6HQ67oYSp0WE3GOE7Gu0IBIbYeXMkfAGXpVEoyjS8L3XA4bDjRc9xeDydBh/RkfAFlsZXVS0C63R24KizxVB6YLXWsY791O/j5CbEKKAFaTJBJyrRrvtUfnnutiZEuRd4UcLF6FrMJlD2+iDKrga0O6djp7e8M295fy3mX7nE0glzp68XbF07TYgCgNPZgtBwGJIUMRTqczj5Zq3xims9T1OTijKNYGAor78vp7M5Yw0iIQlclkbXtHHq9w9CVWMGwad/ThSDhjo/KkgJSaCt7TgmlWkEAkOoEQQofzONX9y7x3wlNwOtOZxUpnCyp8sgOhVlGi80tSb9TAU4HJpNldPZDEISeKGpFe52Fzsf6Os18+EC4m53wec7heFQ2OBKIEkRNttdny6nUcHQcDhNMOaC/vgZONMHt3v17/hk0pKPxBNpzVWAJh4zHUOp+0yLel4BsH53gEJfS5aetGPhkAeLq8yo50K0eOFp+tWYTWh1JFcDWHrCjvlXLm2bRcTygTosHPKwkWiFYCcKUWAlAkEL+TmcQiH9eURrEkmKC1mOGtKgnhc7cbDhSMaGDkISaGpuhd8/CEkaxdlkKjSfEJIwrFtVYzjYcNiQ4s6UplfVGEKhMBNSekQxCH+yiZE1MOqEFu3O9no74G53weGw46tfbcbt2z/JKVOhCdAW+FPKD6jYBIBrE2/hqLMZkjTKoo+Tk1NQlCkmREPDYfT7ByEIAi5L+au9P+pshjweZdFQQRAgSaNsHXpRp98Pm40IZkL/HTqdzes6Z1ORni83haK5TlSZsXDs9LY1+3Jyg4vRbCSFaOnkKBYOeTD/yqVtN6RfOHYaSwfqUX6x12ignwdo1/xOE6KAdjIUBIF1aXI4hULrHCZM8FHxCaw0sdTUmDP+fjo8nYjHEyxamItQk+UofN2n0hr3+v2DhjpNf1KE6QXQxER6ZuV3fqc76ftojCrRxhtBEFCTbFKhEUjK8HAYgiDgZa9mYC4IZvz0p3dhMpk2FQGUpAiefuY5AMBI+DwGzvRBkkZZXXgoFE7WX3ZAEMxM/JEUs31Fmdb2k+8ULJZauN1a41C+5q/TfaUoU7BYajEwQCPF2n7Uf7/Or1Lfz1F0eDohisG82+NxSy3OToaL0SyUX+xlQrSYipmpZxqtX80HtAOSFrLvJCFKsSVFQr4uNBxOJvQdwgAQjyfYcUePPctj6SJTkiKQpFG43S6oagyCILBlrRdVjaHNdQIeTyeamlshigH4unsNdYq0dhGg6fkIHDoPSFEM4kc/upXcdsK27c3/dBW2DE01VPgRQpLRTw8GBvoM61DVGGzWOrYOWY7ir/7q+ygrK9vUuSQ0HEZcZ7zu9XYgEBhiwpt+RoulFk3NrWnv19LP9WxEJbWlo9Hc/InRZiY46fQhr9cDQRDgfclYLuRw2DEWuZTsbh+Fr7sXTc1ac5Y+0r5R6I04h7PT4WI0A6VXgyidCBedEAW0mpfFBhczyM8VOlkJ0IrSd2oDkNWmRZu4GOUUEmrafVmKsBnsACCeC7L0d6bfUCgZPVQmNSG1mXpReTwKSRrVmmYstbjx3vU06x798U87uKkIo9N0nnnmKfZaRZmGr7sXAPAl+29kXCegpeVHwhcwcOa0QXQODJ5lr/N1n4LHo/n+mkwm/PEfn9vQ56PUCGYQQpL2QhFDvaqqxtCYjLZ6kpFmGpHUusA1wXpt4gp6eroQCAwxWzr6veibiXKlp6fLID4HzpzG7Y/ezxgRdjjsuPHeddx47zquTbyFkfAFjIQv4GRPV07OBjv1nM3h6OENTCmYZqZR9vqg1qyUpWtvu7l/rI+Z/653QhMhCUM3JKBdjGjtVb6mgeTKWh292aCvV5TpojDk5+xeaMOc3k4sFAqzVPhjGcQBFYaTyjQGBk5vytZHb1K/2vQaKkj1wlSLqmq+yP/y9/4PHD7y26wOkxCCI0eexx/+4R9j//6HDCl9uo44SWBSmcKkMsW62Klf5Jkzp9HdfYpFiwVBwJtjlzZ9PhkY6MOkMg1RDBiWSbu+3RYXSDyBOCGse1tVYxhOcRLQfw5CEiyCLEmRvNVtuttdazaGpbISTc3P+dZmrYOiTO3IjBaHQ+FiVM9sAuUXtSjBfffQtteIZqXKjPtf62NGvffda3eH0gsntTYhJMHsPeiIz0KjqjGI54Jao0SNGT1JDz4KtYdR1Rjr1ORwig2anqUNc3QMI40wppqZq2qM1REOnDmtdTMnBZ08Hk2zSVoPFkstfN2noKoxg6+kvutZHwn0dfciHk8gGBiC3X4QA2f6IJ4LQBAEuNu1cZGKMp1mDXWypwskTpIOACslADRSSs8bL3s9zOYoV/skQTDj2sQVSFKEjS1NXWaqa8bJnq6MApOO5dTfOGRq0trJ9PR0oTHpt8zh7FRMs/duL2/3RhQLZa8PouyNQSx8uatoo6J6Kk89B9xLYP6bV9a0rHj6mecwEr7A7sZ9yUhGT5aTeCHwJH35rNZ6VvR/473rac/T8Xkbod8/CL9/cEs/D2fvIstRNDW3sihgaDgMUQzA0WhnvooUQhJ45NEntSaXM32IJyeGaR3fnnWLUbpOPU5nS9aUP10voEUWvd6OTf82aEofgGHyTjFDS5AUZQpWaz2CgSE2a11/3uFwONsPj4wmMc1Mo/RqEEtP2LHwm57t3px1sfDlLq3jfyK8qnim49v0TQyiGIC73bVlwk0fdaERFr0HIX2eNgJsFt5RytkKHA47rk28BUBLtw6cOZ31tyQIZhY9palyQDMo38jvb2UCjpaidzTa4Xa7skYhtQjjW5hUpmCz1uckIFfzTi1GtKlIx1n3/8BAHwTBzBwBOBxOccHFaBKanl/4SlfxpudTWGxwabN0rwZXFaPxFMsT8VwgaUXSl+Ud+afNdRwWSy3GIpfQ5joBVY0ZOk4vJ1Oetk1eMGk9Xaq9C4dTKFLF3Wqp6bGxS5CkiKEhaKOlMXT5Nmvduo3bs5mY73YmJ6egqjE4nS2GfSXUCAbjfg6HUxxwMQqte77kZpTNvN1J6KOj2ZqZ9Ok1atIdyPP0l9WgkVl3uwu+7lMsPalvNGKNSwWYl83hFANOZ3NOtdm2pGPEZSmC/Nvl7y6yORvQlD2Hwyku9ry1k+luDGWvD2J5f+2OSc/roaPPyi6eyvoa9Vb6zPZUj0MqGAsBXbc2snM0Y3oy1+gNHR3IxSxnt0I9JfWm+5zMkHj2/bPbGpg4nN3AnhejZX96CqZZUtzd86tRZdbGhM6SrL6j1GQb0KIrjka7wSvQ4+nEC02tEM8VZsyoIGid81o3cZ/Bq5BCO5A3G629lRTS3N6Es5uhN3aTk1NrvHJvQ/eT/gabdefbeGSUwyk29nSavmQygtLJUSw2uHZcel7Pwle6UHY1gNKrwYypekWZYlNhBMGMsbFLbJygvtO0kPZOazVqWK11uDbx1qabJOLJSFENn0bC2cU4HHYoyhT3010Dum9oWZLFUssaKOloTg6HUzzs3cjobAJlbwxiuUrAwpd3uBVQlVmbyjQzBdNM+nx2ejKm3oYHG46gqbkVqhrDwJk+vDl2aUt8RtfCqhspuFFo45JQwyOjnN0LnaaUzylCuxE6KYsQAo+nE/3+QUjSKByNdi7iOZwiZM+K0dKJMEpmprB4qGNNj86dABXUZW8MGh6ntWWTyjReaGplIrSnpwvv37gOr7djV6S2eWSUsxdwOpvhaLTvyQ75jaLZOQmQx6Pw+wdhsdSu24WAw+FsLXszTT+b2NFNS5lY3l+LpQP1KJ0cxcLdGBPYgmBmJvOCIBimwOwmuF0LZ6+QaqrPyYwgmNlAAgDwvtSx6857HM5uYU9OYKKTlu4f68PioY6137BDKLkZRcWrrVhscKWNCJXlaM5j+ooVVY3h6WeeAwDceO86v+BwOBwOx8hsQrNA3EXX/N3EnkvTm+7G2KSlbL6cO5WlJ+xY3l+L0okwTHeNNk0Oh31XClHA2DHLhSiHw+FwUin7bgDlF3tRcpPXWxcje06Mloc6YZolO2rS0kbIVju6m6EG19zMmsPhcNYPHYiyUyi5GUXFt46gsuspVLzamrFhNyOzCZReLYx1ISc/7CkxWjIZQcnNKBZtLTvaymk1FhtcWaOjuxX9hCkOh8PhrI8213F4PJ3bvRnrovRqUBOgd2Ms1V4ysz6/3bLvBmCa1foKlh/i14liZO+I0d1k5bQGey06KiXn2vMOYw5nazjYcIRPgdrhqGoMqhpDaDhcsOl7+aLkZpQ1Hc9/8woWvtKFha+sb3w3Lc3jFDd7RowarJwO7G7Rspujo6knTUIS7DHqwcjhcAqLokwVvYDhrA9Hox1+fxEHLmYTKH/thDYp8Wt9zCmG9kisBZ2ySFmu5vZ/xcieEKM7ff78Zrj/tT4A2g9xt9DmOpGcGrVSJyTLWjG6IAjczJrD2QK4CN1dOJ0tiOtu6osNmmJftLVg+UAdKl5tRXngBMounkLZ64MovRrMGngpvRrUpizaWrBcxUVoMbMnfEbL3hjU7qqOnd6VTUuZWLI1Y9HWov0Qb0Z3fI2sokxDkkZx+6MPDK4A1EOQC1EOZ2ug6fl4nHv77mRqajRxFicEI+HzxVl2MZtA6TXtHL94qAPL1QJMH8eMjUvVZiY0F/evOOSYZqZZEOq+ewiVvc9t6aZzNsauF6MlkxHNW8zWsuusnNZi4WunUTo5ivJQJ+ZOX9/uzcmJSSVzobqSfNzdrn23shyFr/sUbNY6Pm2Fw+FwsiAIZgiCwDJNxWj9Z7obg+luDEtP2FlAZe6bb2lPrhJYMt2NoSJwHAA0z+09EoTayez6NH3Z1cCeaFrKxPL+WizaWmC6G0PJZGS7Nycv6FNJkhSBqsYgCAI+97nH4Os+hTbXCahqDE5nyzZuJYez+6GWapzNIYrBorBWUrLc6BcD1BN06YDOtq/KDFSZYZqZRsW3jqRd20x3Y6gYagXuJXDfPbTjs4J7hV0tRkuvBjUrpwbXrm9aysbC104D0PxVdzIknp4alMe1E1Vd3dM49vV/DFEMwGatw7WJK3A6m7dlOzmcvQL9TXI2h8Nh3/Y6TYulFvEM3yMhCfT7B7d9+6h1U6qgNM1Mo+LVVpTMTGV0jVm0tmD+m1ewZEu/Duy2pt7dwu4Vo3T+fJWgGdzvUZb312KxwQXTLEHpRDivyyZzJqiJ0oz/FQoaSSAkge9853UAQDR6Daoag9frwcjIBe43yuEUkK1K52p14saolyRF0OY6UfB1E5JgmZdCYbXWbfqmmZAEZDmac3TVYqkFIcRQL0pIAh2eTohiMKN7yXpI3SaPp3NzdmCz6a+nY68pqeJyeX8tFo6dTu+0r+ap+mJm19aM0g68hS/vzklLG+H+sT6UToRRdvFUTnWzyp0yyDMVkH5YifhcCdREKcicKevrLeZFOB+fg/fX78FiXtz0evXEiRYZ/df/+lXEYj8GAJgffBCD/2c/qxvlcLYSKgys1ro9dSNEf4t6ZDkKRZmGw2HP2fc3Hidocx1HT08XTvZ0MZFECIGqxgq6rzs8ncyp4/ZH7xdsPZtBlqPwvNhpEIqpjZ3rhe5DVY2x7+uyNApJGkVPT5ehMVRRpvFCkyYCjzqbEScJ1vjk8/WyGn1FmcbBhsMYi1yCw2FHaDjMGk1DoTC83vXPhqcG9aXXwkC1mfWAAMDCsdOaZePNKEx3Y2vaPC1XCTAhBtMswfK6t4CzVexKMWqamdbmzx+o3zNWTqtSZcbCIQ/KrgZQHjiB+57z634rmTPh7Pc/BenDSih3Vg4XoXIZFvMiLOZF1FQupb1PTZRCnqmA+E41xHeq4TgwD+fn5+D94r1NfQShRjvRKso0fvd3T+PfvPpvtY9WtQ9/9Edn856WJ3MmCJX8lMVZHSqQJGkUgiAUnXApBLQL+1ZK1KzNdYINoMjHvnA47Ojp6YLfP8giaoQQBAJDaUJUkiJ5uxmQpAgkaRQDZ/o2vQxCEmnCq98/CEejnQk8QhKIx4lhm0UxCKu1jqXwfd2n4G53sfObLEdZZLinpwuKMp1T9JYK2EllClZrHVQ1huGQJvZsuvHK9Di3WGphs9bhshQBIQS+7lNwfrUZoeEwBgb6IAhmJjwVZRoWSy18vhV7wY3Wpy58pQulE2GUTo6idFI7tparBK0W1NYM08w0Sm5GUaJE2FQmiuluDCWKFlk33SMw3dNunsou/h6WDzyrRVRnE5rTDq8t3XZ2pRgtv9ibtHI6v+ejopSFr3ShVBldl9UTmTNh8k45hqf34fKH+5gwcxyYR3vdL2F7+D6sDy+sa73Sh5UQv18NeaYC8kwFfG8/CO8X7204WlojaBdAVY3h7bfHAQBPPvk4Ln/nT/IeIRHfqUb/9x6AxbyIN7/+MRelnKycFQNMiBYqUqco0wgNhyEIZhx1tuR10hgVM9lu5kLD4WQqeAperwcDZ06zLuy4LuXa7x9k0bTh4TCsOiGTCyeTgksUA6isrIDT2ZKWASEkAVEMwGKpzYuDhravBbjdrrRoo6JM47I0ys5HVDjS5+h3c1YMwO8fZM/LchR+/yAUZwscDrvhJmYkfIHt/37/IE72dMFmq0eb6wQUZQqyHGXCnprTj4TPQzwXhCxHcW3iyqZLJ6jgpMI3FAqDJCPe+mVqEe8ptq09agxPP/OcQQhPTk7BZqtHKClm1VvaMukNhMfTictSBIGNbGCVGXOD72v9HzNTWH4o6RWevK4v2ZqBqwGtNyRFjJa9MZixNK1k5m+Amb9hf3MRWhzsOjHKmpZ28fz5TVFlxn33kGYY/NoJzA2mRy0yRUFpRNP5+NymUu3Ox7X3qolShKarWKRUfKcazsfnMPD8JxtaLr2ztlrrMRI+XxAh6nv7QTgOzGPyTjkuf7gP7mdns76ezJkQnyvJWxkCZ+egiaAgLJZajEUuadGw4XDey0XaXMchCAJqBDOGh8MYi1xKO+4VZRodHi11OzBw2rANijKNSWWKPUZT3DTtSsVC6naLYhC+7l44Gu1wOlsgigGQ+EpkkgoRuh+czhYoyjTi8QS8L60/FbsWwcAQbF/4a/zkJ3+H/Q/9StrzgmDG2NilvK1PUaZgsdSmCTwqIKn4pMKQpqNp5NPtdmFYFx10ONInHFHxDsCwHwkhUG/F0OHphKJMwWqtZ9OuamoEyONROBrtLDr65lj6sbAR6Gek2+dudyFOEmzbKLRZlIpt/d82myZopT+PYHh4RcyKYoAt093uYjc1mymxWDzUgUxn2KUn7Lh/rC/jvPn77iEsWltQMjOleZPe+RFMib/D0mcex/LDfw+oMjMjfc72s7vEqL5paQ9aOa3F0hP2jOl6eaYiLQrqfHwOJ7/083VHQNfCYl7EyS/9HCe/9HNNjH6/GtKHlZA+rIT3i/cw8Pwnq75fX7tktdbjzbFLeW+kIHMmFhEdORrHC99+aNXXq4lStF2uYcK9p+EXOPmln+d1mzjFiyxHQQiB19sBi6UWkhSBx9MJEk9sqC5uLeLxBARByCq4CEmwWj632wWPpxM1gmBI7YrnAkwQ+Lp7MRa5ZBBIqaleWY6i3z+o3fSNXIAgmA2RrRrBjMlkk8plaRSEEFZjGQwM5X0Ixey9WZSUlODC//UfYP/vD6YJ59QU+Gah89rd7enLUdVYMjqo3QQfdbbgYMNh+LpP4drEFdQIZsjjUYNQjxMCSYow8QZoIp9GclU1llZ7S0Wc1+uB1VoHj6cT4rkgLI9pgosuy93uYqKVCuKNQssfBEFAMDAEp7MZvm4trU6FNACQpIsJ3T8+3ylt6l2jHYJghsVSy7ab2upJ0iiczhYMDGjlDsHAEA42HIbH05nXm4fUiKieJVtzxo56TvGxq7rpy17XJi3thfnzm2Xh2GksHahH6eQo/uLtH+Dg8H40fftXEHq3ChbzIgae/wTvf+MORo7G8yZEU/F+8R5ufOOnCDQlYDEvQnynGlX/5jMIvVuV9T2CYEYgMISenq6CCFEAOPv9T4HMmdD+7C8hVC6jpnIpYz0sRfqwEraH7yPQlID14QX4Jz5VUCcBTnFBBzHQVCetlcvU2JMLXm8HFGUKBxuOsHXoOSsGmCg+2dNl2DZAEy/xeILZ9QCA58VOyONRnOzpMhifU9RbMRBCYLPWsd+aUCOwzmurrZ79P7V4KpSt2lkxgMQnn+Af/cOva9uuay6iaBG5wlpNpU6cslrrIAiCQchrHe4riejxpKgXkml9Wfd34DVj7SsVddqy63GypwtHnS1M6Pm6e9lyABiOhc04DEhSBB0ezfLPYqlN+970xzEVmE3NrWhq1qLpJ3UNToHXhrSbgUY7goEhjITP49rEWxgJn2fHj9Vah5HwBcjj0aLwV+UUF7tGjJbcjKLsakCL/vGmpVX562Nh/E+P/r/4rXeaoNwpg+PAPK6138W19rvwfvHeltVIup+dxY1v/JQ1NXnGzDg4vD+rKHW3u5IXz/wLUTJnwvC7+yBULuPlX/8FAKDxwP1V94XFvIj2ul/C/ewsi4iGprMLas7ugjbwWK11rKMeSLe1yQdaGQBhFjnU8oimyAVBwMteD84mhdAtXeqXRnB9vl6WQtW6p+vhdrtw1NmctbFESDYrkWTqVhAECIIZjkZNhExOTrHmQq/XU5C6WfpZe3q6WE0oHW5BicdXZqsryvSm7Z/YqNMUqyNVja2ko5PboyjTTLADK2KdptIBJIXXFI46m2G11rP9T6PHI+ELeNm7cr0aCV9AT08XE3GCYMZY5BL77O/fuI7bH32Ascgl3HjvOq5NXIGj0c4cBtaLqsaYEKV/U2gqXi/unc5mLfqaFJyBwJAh+u9w2DE2dgljukBBptpmp7MZN967vqdcJzjrY3ek6WcTKLuopRYWvsKtnLJB09Chd38V5FPP4vnZt/F7C/8ODV//g23droHnP4G7bhYdYwKUO2XwjJnhGTNvqtFps8gzFVATpQYxvlbK3fn4HPt/GkW+ldg193ecdVJTI7DoJIC0qF0uKMo0a3wJvDYE6c8jCIXCaHMdRyAwhBpBE0A0ckVrFWk6XS84QsnGIkWZgiAIrBnpMUst4pIxqkgbdEicQFGmmQF6TzLy6nDYYbHUQh6Psshwv/8tEGUAACAASURBVH/QkI7Ox42jokwzoSyPR+Fud7Fud79/0NCwJI9H4Xa7WNf3ZqDvk+UoPJ5OTCYbvGzWOoyNXYLT2QJfdy8TmcBK1ND51WaIYkDbtwN9rC7U3e5inea0wzybYLNa69Ies1hqWcSbYihbstVDHo9CkiLrKg8hJKGVkxCCkfAF9PsHoShTICQBQTCzEoijKVPsHA47HI7c0+tciHIysSuunGXfDaBkZgoLhzy8aSkL8kwFXvj2QxDfqQagCcCL+8/h+Z+MGAyEtwvrwwu41n4XY1//GWsYEt+pxtP//tPwvf1gwdPf0g8rAQCO2vlNvX81v1XO7oRGDan5uMVSy0zE82WWromTekjSKJqaW7UmImKs3wMAR6OdCUa6DTSdHUwKNipAR8IX8ObYSo3hy15PWr0h/Ts0HMbBhsOQpFF4vR4migTBjMBrQ6w21dFoh6JMQRQDCIXCefv8mnDTahOpONJqG1sMHftHnc2QpFG80NQKRdestVEEwYyBM30ghCA0HEaNYMZI+DxGRi6wdbvbXZBlrfRh4EwfE4AOhxYxfHPsUjIlfR433ruOQGDIID7zkdnRRy1pFHa95SGyHIU8HoXX64HT2Yya5PbovzN3e7qTAIdTSEyz927vaN8aOhYM1WbMffMtHhVNYSUaWgUyZ4L72VkMPP8Ji/5VfOsISmamsNjgwn137rYo+UK5Uwbx+9WGlH2+m6r0tF2ugfRhJW5846csEkstrqif6mqE3q2CZ8yMkaNxQ8SUs3tRk/Y2lIEzfRBqtEYfR6M9r00akhTBpC565Wi0w2arT/pOat32NEI6cOY0swWiYomWDmzEFqqq+hEAWuSPdomvJlComNmOyJfeKkkficx1mcUoyOhnVdUYrk1cgSxH0dTcCqezBSPhtT2km5paMalM4/0b11ljWijp0pDvxjMOZ73s7DT9bIJ5is6/yD1FU1HulLHUt/XhBZz80s/ThNL8K5dQ8WorSifCWHrCntOEpnxifXgBgaYEAk0JZrdEu++p3+lqlksbhUY2acNS//cegPhONXt84PlPshr2kzkTxO9XQ6hcLljTF6f4sFhqcW3iLfh8vXA6W+D1dmg1nNYg5PEomppa8yZInc7mjI1BTmcz3O0uXJYiBhF2beKKwfdyM96kVOAGdZG91djO9KuQjGDmczJTMQtRfcR7o593MqVUgEb41Vt8Zjtn+9jRYrTs9UGU3Izy9HwGqHF7pmiogSoz7nvOo+JbR1Ae0urLikWQUrxfvAfvF+8xSyhqoE/rSp2fn4PjwObS65TUyKdypwxHH/8lM/lfrZGp/3sPQLlThp6GX3C/0T2GNVlLSKFCUBSDhvrJQhIIDGU0Es/VHF/zIiWGUZHFzm6vR/T5etlwgdSbk/WWRhBCUCOsfJ9UlFJXBA5nO9ixNaOlV4Na9/yBeq1piQNAi9K1Xa6B7+0HAYBFF1cTU8v7azH/inZBLQ91ouTm1lxENwq1hBp4/hMmPsV3qtH07V/BI//2V+F7+0HIMxWbWjbdP5N3ygEAI0fjGHj+E8gxrda26dvpZtt0/eI71bA+vABH7TwTypy9jdfbsa6UaTFDu8Tz2ZDF2Twrk7iEtIYmAKz2cy0sllrmEasn35ZkHM5G2JGRUdPMNDO3v+/h6XmKcqcMbZdroCZK4Tgwz3w818PygTqWsq94tRXzr1wq2mgzjZSqiVI2blRNlDJhaDEvwvn4HNx1s+tOmztq5yG+U61FW3/9HuK/LDGk6QGtLlRfGuB7+0HWEKbcKUsTrPr6Uw5np0HTt1ykFAfU59SWEqXeqL+qo9GO0HBY1+zGI6Kc7WfHRUZNd2Os+/u+ewjL+3d3Wma90G55ak80cjS+YSG09ISdRUgrXm0t2ggpxWJeZNHSsa//jKXJqTA9OLwfB4f3ryti6nx8Du5nZ6EmSuF7+0H4Jz4FQKsVDTRpJ2u9MFXulDEhqkeoXIbjwDwGnv9kVcN8DqfYoV3at/LUGc/JDTYtqUYwPL7RxjFqRUUncNHpSru9xIFT3OysyOhsAuWhTphmCe4f6+NjvpKE3q0ypOVzaeyhgnQnREj1OA7Mw3FgHie/9HPIMxWQfliJ0LtVUO6UGYTjajWmgaYEer70C8gzFbCYF1mtqDxTgZ6GXxj2q/XhBYwcjUNNlML68AJqKpdgMS9u2cAADqfQ0M7qVPHD2R6sWcomJlP8TteCesRS31la22zTWWVxOFvNzrF2SgrR0slRLBzyYOHY6e3eoqKANioBQLCJ5M1WqORmdCUCfaxv1fm/xYz0YSUm/64cw+/uS/MqdT87i8ba+3ntyudwdhOqGkNNjVCUneV7kUcefQqEEDbFiJAEDjYchqrGNjTZiNpBUdztLsMAAQ5nqyk92dP1L7d7I9ZD2Xf8KJsIY7HBhYV/cGa7N6coKJQQBbSmpqUn7SidCKP03avAbAJLzx7K2/K3iqceWoSjdh7eX78HR+19WMxLIHMl+Mm9Eih3yiF9WAn/xANQ7pTjJ/dK8JlPLaGGRzc5HADadKl9+yq3ezM4ST744IfJKU6P4Td+47/FH/3R/41vX/wzOJ0t6Pin//O6l2Ox1OKoswX79u3Db/3WIZw86ePfM2db2RGR0dKrQZRf7MXSgXqtppE3LDHvTaFyOe9CVA8dKmCaJVg6UI/7nvO7ok6XNj9JP6xMqyd1HJhH44H7OPr5X3LfUA6HUzTQiKYgCHC7XQiFtPGvdOoTh7NTKXoxWjIZ0fwv+YQlhl6Ivvn1jwsvmHQlEgB2ZZlE6N0qSD/UTPX10M78fHiZcjgcTq60uU5AkrRzsSAIGBg4venxpxxOsVDUYpSN+oQ2KWj5AL/z6//eA/BPfGrrhKgOGqEGtDT+wpe7is4gPx/QiKl+FCmF1pkeffyXvFmJw+FsC6IYBKBN4OJd8JzdQPGK0dmEZi80M4V5z4U93zlP5kzwvf0gQu9WsU7ubfGwTI5gLZ3Q0kNLT9h3tcUWFaaXP9xnsHYCtHQ+Nbt3HJjn4pTD4XA4nE1QnGKUd84bSBWib379420XPiU3o5rN1l3N427hkEebhLWLyyjkmQrIsQqMz5Rn9C2lHqNUoK41RpTD4XA4HE6RitGyi6dQdjXAhSg0IdoxJkD6sBK/8ch9XG79WVEJHH3qHtBE6WKDa0+UVEgfVkKOVUC5U5bVVF8vUC3CIizmRV57yuFwOByOjqITo7xzfgXlThla/6wGP/655o859JsJeGxF6Ik5m0DZdwMoe2OQPbT0hB2LtpYd60+6GaQPK6EmSqH8XRkm75RDuZN9poTFvKgJVPMiLIL2/zySyuFwdjqqGoPfP5jVtzQ0HEaNIMDpLHzpHfVhvfHe9YKvi5MbRSVGuRDVUO6U4V/ID+LqLS3aVlW2jKHDn+wIc/bSq0GUXQ2w9D0ATZQedO3Jul/pw0rE50qgklKMz5Rj8k55Wu1pKkLlMhslKlQuw2JeRE3lEh4zL8H2q/eZkOVwOJxiIzQchsfTibHIJTbFS4+v+xQAYOBM4bOeijKNgw2HMXvvdsHXxcmNohkHWno1iLLXB7FcJeC+e2hPClEyZ8LlD/fhn199EJ/Ma4LlV6uXIP+Dj7enWWkTLB7qwOKhDpTcjGqG+RNhlE6OMluoxQYXFq0te0aYZvN/TRWpaqKUTYgicyaQuZVpUdkirM7H55hIBQCLoB0j9FixmBd3zHHD4ex2ZDkKQhKGiCAhCYRCYXi9xZFBOthwBANnTmcUkeuFxBOwWGqzLoPECeIksenlbwQ6KpVT/BSFGKUR0eX9tZjvvLRrO7OzQRtjLkztYyl5APjNx+bxH74a35Gp26Un7KzTvvRqEKWTowaBCmgR0+Vfq9P+3QM1pnqyiVS9II3PlbB/VVKKW4kSqIlSVp+a6om6GhbzIouyAmAiltexcjhbg3gumJaaPisGcEuNgZBEUYxcVZQpKMp0TmJUvRVb+0VbBIlvjejl5M62ilHTzDTK3hhE6eTorprusxZkzoTJO+WQYxWQPqxMi3ztK13GnzgJmj5XmKlKWw2NlgIwCtPJUWByFGVvaBHxpSftWP61OixXC1h+qBao1k7Oy1UCW5ZplgD3Eiv/D03U7pZI+nojmfrI6q1EiUG46v+lUJG7njpWgItVDiffKMoUHI1GkTc8HEZPT1fehaiiTGNSmdqUGX6ckJzWTeJrv389r8kHuX4WztZRODE6m4Dpbiw94jWbQMnMFEomIyidCMM0S/aULRBtctELBRMAGvus+/QC3mrbfuumQqEXpiWTERYtNc0SJk43SokSwX3P+XxvalGz3vGv9FhTE6WIz5VojxFNmI7PlCM+V8IEqr5UINs6HbXzK41XvARg19PvH8TJnq7t3oxdQTxDlE5VY6gRhAyvzo3QcBiiGMBRZ8u6hS5Jps5Jjin0OElk/Kypr9kKbqnFE6XlrE5BxGh//wC+eDeKr81/DwAwa1qxvalaXomw3Cx5FH+67zBmri0D1wYKsSnbyk/La/H/7fsC/nbfF/DT8s+wxyuWZvGpZU1MLJtKMWeqQsnyAg5+8h3Y/utb6FO2a4u3CVMbHqr6BAeW7uLA0scAgIeWP8H+pU8AAFVYOWZmUYFZ00p6+p6pAj9+L46rvl5wNsZ/k/y3EcAnpfvx0/JazJVUI1G6H0g+9knpp/HjyicBJAcApJQGOB+fw8jR+BZuNWerICQBv38QL3s9RZFC3ukQQgyROnULhJKqxjY8sz7fUUtZjsJiqTVMiiJbFLHcKtHLyZ2CiNF/94d/jI8//hleeRaw7Qc++4AmJuJJTTF5F/izW8APPv4IwKVCbEJu9P4M2FcD/OxHwEc/AD6aBOI/0v6m/6ayr0b799EvAJ97Hnj2t4GHv7Dy/Ec/AP72beC/+wbmKx7EPHSjJv/2bSy+/s/wFx/9AH9R2E+2i/nedm/A7ubZvw987n/EF5rd+BlqWP2pu674HR44ubEZQbNXkeVoxnrLTFHH716VAaAgQn8zgjKeJxGaum5CEggNhw0R9rUip7kiy1HI49EtKwfg5E5BxOiPZ97Dz3/+i6zP/w8AXirEivMEmY8j+l/n8Ia6H6M/Ogry7N/f1HLq98/jy5ZfoOWzv0D9fjOAowDu4LWpeUzdrYBQsYSWz95D/f7PQvhXf5rXz8DhFIIHHlgA8NPt3oxtRZajsNnq90y0MF8iZS/g9w/C4cgeYCFxAkIS6PcP4tvf/jMAQE1N/tP0a0UEFWUa8TjJKJxzjSamvt/hsMPX3btl5R6iGES/X/O8/vSnH9qSdXJyp2A1ow888KlCLbrgPADg6w8BX6+/B+Ae5JkKrZ5O19EMgNXg6bGYF+H8/BxsD9/XeUGWJ//T+OcNCwDocyWAPkrK4XCKGr9/EE5nS9HY8RQKKrZz7a7eSzidLas+fys2gw5PJ2Q5Ck/HP8bA4B8UZDtoRDDbDZPFUot+/+CWfK+CYMbAmT7j9unS9DSKmUmsqmrMkN7Xv8fvH4TVVg/vSx3sNaHhMHzdvfB6PXC3u/CNb3gBAK8FLuBFz/F8fixOnikKa6dih3cSczgcinortme6dAVBKLrP2u8fhKJMYyRcfE2L2W5QEgktWhiN/hWqq6sRDAwVdL+uZa8kCOZ17T9FmYbP14uBgb60Ug1JikAUA2h3u9K69gkhBrsqvaWVkIwE0+f9/kFMKtMGMUqjx6IYAAAMnOlj+1ZRptHmOgFCCOTxKEKhMG5/9D4ALSoqCAJOJh0KzMn1C+a9kcXYyaSH9jgcDoezKnupS1dRprd7ExiKMg2/fxCSNApRDG735oCQxJrd57IcxeEjvw0AeOihX8H7N67D6WzO6IEpikFIUiTn7aLNUZmiimtBo6qEJPBCUyvk8Whas5UoBtHmOg55XItQrrYNAPD0M89lfI2iTGu1nSnC/KwYQCik2V45GrU0v5r0Y+3wdKKmxoxAYAhOZwsIIVDVGCQpAkWZgtvtYiKYfpY7d/Z2adFOgItRDicLuVqccHYve6VL12KphbKJKTayHM1ZVDU1taKq+hGDGL4srVi/0eifNvLxSE7ryobH05lV9NK55y80ta66jDbXCbjatNd85jO/mjV1LopB+Lp70eHpzGmbcxGies6KARBC4Gi0GyKbqhpjNZkAYLXWG95Xk/x8tNb45d/pzugcoKoxw/dJEcUg/P5B2Kx1ONnThXa3FnVVlGmEQmEoyhTa29OjsXQdNIJLSIJFnz//+b+3sQ/P2XK4GOVwstDh6TScdDkcylbY8hQL+s7n9dyg9fsH0dTcijbXcXhyEFZU8OvFHhWmjkY7vC+tpMQLWb+rPwfoxWkoFIaqxtasu7z90fvo6no57fFUmye6HkIIZDm66e2l35Gwhn8pIQm0uU5k2K4EVDXGPmfnK/+r4XnxXJBFMgVBQDAwZHjeatPEqaJMQ1Vj+I//0dicS8X4pDJlEPqEJAzNR1RM639r9MbIZq1HaDgMWY7Caq2HxVLL9qckRSBJEXR4OvHjH38EAPhn//s3ERoOr7o/ONsLF6McThYUZSrn6AJnd7KTxKiv+1Ta9vb7B+HrPrXme2sEsyGFmnqD5us+hUcefYr9TaNalFz205tjl2Cx1IIQwoQEFWmNDjv7bVqtdZuaNLQeGh12tn5ZjrIGGVWNQR6PsvXrWe0GVr8v9cLe130KhBAmIOmyNwPd5zVruD3E4wSSNJr2HRFC4PcPghCC2tpfw+KCcbiFlIxmUiHKUuLJkgU6ZUo8FzAcY/Tz2pKRVLoO+j12eDrh6+7F0WQUVqgRoCjTrA7U4bCzetN+/yB8Pm3ZA2dOA0DSD1eAJI2izXUckjQKV1srnM4WqGoMHk8nDjYcyUnocwoHF6McTgYkKaJ1cj7GxSjHyFoRp2IjFAobUuaSFIE/2RxCHyckAV/3qTSBqm82AbQbNFovS0gCoVCYiQnadCIIAkvdNubQrS0IZtad7k82LVExlzpWs1AcTa5fq0dcKRfQe1jqJyjRQQHZyHTsdHg6IUmjsFrrWZRRFIObLhPKlKZXlGmEho3HQU3Kd0v/VtUYQsNhCIKAhx/+NGsiotBI+cDAaVitdZDlKHzdp3Cw4TCeevo5OBx2WK31yTrOUbzwwiHD+x3JGwlVjcHRaGed9pI0ioEzfRgY6EvugwBeaGoFIQQDA6chCGZ4X+qAIAgsUBAMDLHItCCY8f6N6wgEhuD1ehAIDGFo6PcxEj6PscglDJzpQ41ghvTnudfkcvIP76bncDIQGg7Daq3nljacNGoEM5SUbuFsrKRM018nikGot2IsslMobNY6yONRlsqmkS2tAaQ52Z18nIkYd7srLdpHje9VNaaN7Eo+RgiBw6FFJUOhMAghsFrrmVh42evJadtP9nRBFAOGNDaNklFW88zMFUEww2KphSSNGgzUJSnCalb1+0pVYxkFJ12OHvo+RZliUUartQ5er4cJsWsTVza97XSf0cYeyu2PPoAgmNO2if5NjwOvtwM2az3aXMdxsOEIRsLnYbHU4qizGaHhsKEEQxAEHHU2Y2CgD4JgxrWJK+z7etnrwf/y4j8xrOfGe9cNfr0j4QuwWGrZPhk40wdftzZVLxAYYpFvi6UWtz96P+vvShDMGaPkDocdDod919ux7WR4ZLSIUNVYxhoeztZCSCLrJBUOR9BFkABNDGVKzWqNNYfxyKNPpj2vqjH4unshy9GC1yVbbfWG1CRNATsa7VpE1KelnWmESr89+lQvFQC0lpM2n9DsAV0u3S89SXudXNCLCyqibRnS4jSiS5L1jvmERmfp56PilApPvaCLxwmzNUol8NqQQSi5210IBIYwcKYP1yau6ITYaYyELyRLJDYeHRVqzGx7tYiyJkRpFFL/ndx477pBTAdeG4LVWg93uwsvez1wOpsxFrkEp7OZRU4DgSGMhC8kb1rq4fV68ObYJQR0KXtAu5GgFksOhz3tWNA/5nQ2G7bD6+3A7L3buP3R+xnFJRXTnN0Dj4wWEVrh9ShCw+GC1UBx1oZ2kTq/2rz2izl7Dio+aLdwh6eTRQLp71ZVY6zxxulsgSRFDD6KVDCd7OkydCoXAkejHaIYQGg4DEejPRnl1Jo+QsPhZNRUEx6+7l4tCpiM+lLhHY+v1PbRKULDKQ0hKkvfE7gzdDtvlvZ2l6H5hDbI6NdL162VJIxibCx/Y6adX21mqWpNfHWwqGDqDastuW2ZRqjS6JyebPvI6Wze9HFx1NkCyRlhhvFOZzNs1vp1jXR1OOxp0dhM253L9nE4meCR0SKC3tGO8wLrbWWYp+g564CQhCEFqo/IhYa1lPXJni54X+rI2iG9FY1Q9Dgel6Mr3ejJx4ZDWm2g96UOQ61j6ghQeTzK0riTyjTOJlPn9LlUrNZ6SFIE/f5BtLlO5NQ0QmsQ2d8p9aKGaT7j0TUN33NZv8NhZxFBut/0CILmf7mdjY/U0P7axBWMhM9nLLvgcIoNLkaLCFoIn+rbxtk6aOMSv+vnZIOmB2mjB4UKPWpRY7HUwu12sTo88dyKjY1WKydsid2MIJhhtdbjshTBpC5lSw3HHQ47JCnCmlYA7XfQ5jrBUvB6SyXabW2x1LLuZc0PdKUu0dfdizbXcfj9g3npXh4Jn2cp4Uy/TSr2ZTlakAYzKuqo+Lw2cQXXJq5kvGF1t7t4CpnD2SBcjBYR9MSWi60HJzdEMQBBEHiZBCcr1JomTojB/otFCuUoMwsXBHPGCUaCYEYwMARFmcLBhiMFn3Lk9WrRWerrqI+Uad3QvbBa69mISO3vOni9Hejp6WJNVgMDfXA0apFCWuvodLZoTUWNdoyEL8Dr9SSN0lvQ09OF929czznLYLHU4trElYzNXjRS2tSsdV4X4kbSYqlNi3hy2zcOJ3+YZu/dXt7ujeCs8PQzzyEeT7BZu5ytQ4sGHYfT2VKUc685xQFtTBIEAYQQeL0ehEJahPP2R++zSTq0C7gpOVJREIS037UsR9HU3ApBEHBt4krBBA6dFkTrRWldIE2h26x1GBm5AEEww+PpxKQyjTfHLu2ICJ+iTKPD0wlVjRk6ujkczs6BR0aLDDprt5jmQe82FGU64/6lUVF9owmHk8qKr+aK5yU1Z9ePIKwRBIhiUEuFJ9PbNDKpr90ci1xiqe9CIQhmBF7TZnnrj++R8Hnc/uh9jOmEZyAwhGsTV3aMoLNa63Bt4grzmNwp283hcFbg3fRFBu181cac8aLzQiDLUcQJMexfSYpAHo/yYn/Omuj9GKnnpWPcDkWZwlkxgH2VlQDAJvVYrfUYGbmQrGfUhJK+A5/WOD5W4LRvpq7o3QQXoRzOzoVHRosMXjdaeBRlyjA1BViJinpzNOnm7A00v0aB+ShSGzC/fxB/Er4Ed7uLTZih6W6ns5n9vrX6xz7E45ovJvV15HA4nL0IrxktQp5+5jmoagw33rvOi+QLQJvrBByNK9M46DQRr9dT8Gk4nN2LLGu2QkedLTxKx+FwOBuAR0aLkP+/vTsObvs87wT/pSTbkZ3Fi8ZtKjWGkN5dLLnk/dh2q5WKHtFtrTHJJlC7bI/o7JJz0s0SmanhMntDcLZHkTcnEdNbkb0N65+3F2BvpRbsbcH1qnv6pSHpynUXvEWlqpeGvyNr2d6kgeGutE2d4MU2VmzL0v3xw/sSIAAKJEECJL+fmUwsEgReACTw4Hmf53nDzzsZkuJzhKl+imc7qrOkK80MJFoPNYOSgSgR0fowGG1C/f3OWCHzpViDV7I7SSl1k4ka3h0ODzALTURE1AAMRpuQU1/WXXLMHdVfZHhMDyd/IRyClPltGUJOREREKxiMNik1uLn41Baqj1wuj2/99bdgFs6gn7h4AYDT4RyJjPEDABER0TZiMNqkTge6AUAP017Nsub0zEJaHyklZP6/AADC4RCEcOHEyWeRSqURb/C50kRERHsNg9EmJYQL/X1BSCnLto6lzDszDN9mBm+9pMwDAN577y7efOPPADjd9blcHvHYFM+kJ6Jtk8lk0Rs8y0NOaM/jaKcmJmUehw4/XXaM4Hh0EtHopD5ukNbn4OOHAEAf5yiEYCBKRNuu2ms80V7DzGgTq5YdnS78d7vR1qil7WgjRcch+jt8uHH9GgNRItp2QrgwMjIEKSUiw2ONXg5RwzAz2oSkzOO5zh7MJC/B7RYln5zVJ2kAuHP7Tc403AQp83z8aMeRMo8XzRj6+4LbUt/Mv5OtVfyazoNOaK9iZrRJZTJZmC/Fy7KjqZRzTKhhtPENYgOKO+X5+NFOpMp0tqPOUMo8jh47zgkTW0gIlz6GmNNTaK9iMNqEhHBhJnkJphlDb/AsJiac0UPFY4fajdZGLnFHsu1lnDh5qtHLIALgBHrj0Un0Bs/WPN/WmaIRw8jI0LaUlmQyWUgpGYxusXOF0qFq01OIdjsGo03K7/dhJnkZljWL3t4zCIdDkFLit/6P/xMAINyiwSvceRbtpUYvgQjASilONDoJy5pFKDRYU82gacb0IQ3bIZdzTipjt/fWEsIFr9dTcXoK0V7AYLSJBQJdmLh4AamFNL761UV4vR584xvfBMAt5o3IZLLMKFNTuGrNwraX4PV6dEOdacbWnB1sWXNILaTR1xfctr9/91760Hs339CbV78H0ehkQ9dB1AgMRptcODyA/r4g0ukbePzgQf31jz/xRANXtTO9nckyo0xNwbLmAADh50M4NzKEG9dfhWG0ITI8WjULmVpw6sUbMc4tJ2XtF76bxyOxs3gkdnbrFlRn+1+L47HR4/jY84fwSOws9i3OrXzzbh6P/vopfOz5Q6VfrzN10AmPgaa9iMHoDqDmib5+6039tcOHf7Dq5Ts7e9Ab3DlvBBuhhtevR07m4WZGmZqAXSgZUXWfhtGKV+avoL8vWDUQsaxZGEZbQ7qt315HcLTvzTT2L85i/+Kszjbufy2OfW+lt2p5m3bgMkWiuAAAIABJREFUtRha7joB9/7FWTwaO4NHEoN49Is9eGz0OPa94zxfLd/euiBRCBf8HT4AKx9WiPYKBqM7xOoB9ydPHq94OdteRmoh7XTj7+LjQl80YwiFBtf1MzK3juwOUZ0Vf4DK5co/TAnhQqzK4QsqW+b3+7Z0jZXWBDgf5GpVErAddH7+kZdHceDl5p2j+eBg+Y7J/utJ7HsrjZa7Eg8OCtz7uSF89DMDW7qOjsLzq7LgRHvFgUYvgGoXi03BKAy6f1h2ZOLieYQ+Pwi/3wdjF9ZJvhAO4eix44it42dyMg+D2/TUIL29ZwAA8/NX9NdqrclUW/cqc7bd1vNBTmUR7z/lvFa1vNvcW84t7yxj3ztL+Ki9G/c/40PLexK4m8eB15xXlw++cMW5Lwe3flflhXAI0eikHuFHtFfs6WDUsuYwHp1EPDa1YwK2cHjtT+aq+9XtFoh9aQqRyGjJm99OkkqlkVpI61rPiYvn9feEcOF0oAuJ6WTNNXRyPXVvRHWmsl0bqQdUkyC2+3VKBcvryYyqrfn7n/GV/LtZqfKBj04Ecb99JSvdcldi//VCZ/s2BKKA87qmjim27eUd875EtFl7dps+k8liIDSIcHhA/8FLmYdtL2+oHrEZ+f0+PUx5J0mnb+DTP2xgdCzqnP7iFkgkkmXb8n19QdZW0Y6zkW5plRndTL3oZrJt6wmgVSZUBaPNXCsKrGRyHzxVGvh9ZDgNRTog3SZq4gezo7SX7NlgNLWQhpQS3iMeJKaTOHHyFI4eO44TJ5/FocNPo7OzZ0d3NKoM6U47cz2TySL0+S/gU586jKv/97/GxMXzzv8mziMxnSx5Tvx+H2x7qeYPD5Xq9Ii221Vrbt1BpcxJCFF7iUkmky2ZW2qacXR29ax7huVGRkipGsv7T++MYFRlbh88Wfqc6H9vc2bXaHfKG9Y1wYBoh9uzwajS2dWDUGgQbuFCf38QsdgURkaGkFpI78jTenb6XMBodBLnRiL41Kd+qOSN8HSgG0KIsjfTW6/fLLmclHmYZrzyB4kHD6re7m7JhlNjmWYchw4fxfga2U9nC9bJxq3nA6/bXXtgaNvLMM2Y/nuxrFkApTWn6/mdr7XEZd/i3EqNaGFrW3ei323O4Krl3WxZIApg27bmV1OvZzxogPaSPRmMSpnX27uBQDduXH8V8/NXMHHxPPr7gjg3MoSZ5GVIKXdcR7p6Icu8vTOzurHYFH7ohw5V3KKSUpaNmFn9gt3bewaWNVvWjSplHn/73e/ig/ffr3i7vb1nuC1Gm6KO9wyHB8oynyrwCxRmSSq1ngq2rppNQHfdO8d55vXfg1qXOgGqFiojW0vwuv9G6YfFlneWV4LT99b4+bv5hjU6VQuSW94pvLZsc1DqXkcGnGi32JPBaCqV1pkCw2itWCTu3+EjNmSFLemdUg/b3t4GKWXJWtWZzavfzE+cfFb/t2p4Wj0GCwDGo5N45JEDuHv3e2W39+KLMSzay2gvbI8RbUQikYSUEv19wbLfP5UB9Xo9JYFqpb/TStbbfKcaYYp3CdQkDvX9meSldV3nw7K4Le9mse9N5/WyeGbnw7S8s4zHRo/jsbHja9dnbtF2+YODomKgrAJrVTtavI4DL485Q/2bvDmLaKfYk8FoJpPVn/arDXPWXY07bDZltW1604zjuc6eNbcPm4XO7haem8R0EpHhURhGW8mcRdteLnljt+1lCCHKslKpVBqJRBKPPfpYxQzT//7PX8LRp/8bHrFKm6KO96xUD7oSELaWfKCqtS5QCLHumud2oxVSSp3xX30Ubq11q6o8IPeQ18J99pyuFwWcIHP/jSQeHBR4cFCg5a4sz37ezePRLzoZ2o9OBrHPrtyQ2PJuFo+NHscjifXNFq7Fg6day9d2N499b6ZLal/1er90BvuvJ9HybvahzU373kqv+9Qm9TvBAzpoL9mTwSiw8sK83u2vZrcypLr0jcN8yZmZd65w/nGzc7n+Dv7hP/rHiAyPIRIZgxACM8lLJQFj8YcKwClNWF1XJ2UekeExZ+v000fKPlzY9jLu3PnPetg00UblZB7eI5UDPJ0ZPeIpqdustS7QLVzrzo6q32lVN1qcGQVq79ZWf2MPK/1RtZf3Puu8xjz6xR60vJvFRyeDulN9dTPTIy+PouWuxL3PDuHezw05Mz5Xu5vHI4lBtNyVW3Icp+r6P/Bvx1Zur7Cuj35moGSb/sAfxbDvrfTa6y3ySGIQB76yvgSA+p040oCTtogaZU8Go8Lt0kFota0nZ0tbNuTovXpYvR2fyWTh9Xp2TPbv2NGn8fWv/yVMM4Z2oxU3rl8rey6cejhZMt5pdfZoPDqJTCaLF8Ih+P2+whQF5zJS5hGJjGLfvn34xjf+cuvvFO1qtr0EUWVnwrJmIYRAe3tpdr/WBibVYb2eppYXCmPdVLPU6r+f0OcHa6qJVxm6h5UU3G/vwkdGNz46GcT9p9rQclfi/md8uPfZIR3wFWcSVWbx/md8+OhnBvDgSQ8+7J8qu14VAG7U/tfia2ZU1Xr3L87iY88fwmOjTrnA/afacO9nV0bjtbybxf7X4viovdsJUh936cH+leh62XXUnGYyWf0hYa0DDmx7GZ2dPTui7IqoFnsyGHULod8Eqr0ZqMaCnZgxW11eoO7jerZ9Gl1f+vdO/F10dZ3C/NwVzMxcrvihICclMpls1a1O04zDNGOYmDgPIVwIfM4ZcxWJjAJwjhRNLaTR3x/E17/+TZw4eYov7lR36ohev99XKP9xrWQbawxGVWBy1Xp4DaYihKvk72Z1bXzx385aWVJ1HQ8rKbj/GR/u/dJ54KALH/zaNbx//iY++Pxl4KBLB3Ut7yyvnFdfCEw/OrlSX7u6q33fW2nsfy2O+0+1VTyysxYHXouV132u8sEXruDezw3hwZMeJyj+pQv44AtXSgLJ/deTK9nSwv0tHpK/mqqXffB47es2X4pDSllWkrSaYbRiYuLCjkkuED3MngxGSxoIqrzAqmzb6k+nlYIVKfPrnt9XSSaTxbFnjm9qvmkqlYbb7UJiOoljzxxHZ2dP1cHwUuYRCg3qyx58fKWL3TRjeNEsPWwzMjxWcj8T00n0Bs9uWQD33nff02/ga2kvbD8GPtcFKSU6O3tw4uQpRIZHEQ6HdDOJ3++Dv8OHxHQSBx8/hGh0EobRhn/2v/2vuHH9Gvr7guyop7pTAWT4+ZXT01bXbz6MCkzWO91D1adWqqUWQsC2l2Hby+gNnq16HSrbW62+vpoHT3pWgrmDLrz/a6+WBHiqnvSj9iqB4t08DvzBJPC4Cx+GLm1sNFShS3/1QPsyB12499khJ4D+tWtl2/Nqvfefals5WWoNKou6HplMVjdqFp82Vw1PZ6LdZM8Go8VBaKVgyraXyl7Apcyjt/dMSUAmZR4nTj6LUGhQb5uEQoM6oHTm/dX2orR6qPt6RYbH0Bs8q2spDaMNmbeziAw7mcDVW4iJRBKL9jLcQmAmeRlCCP2mJNyiZEvQsuZgmjFEIk5dlWnGEQoNwrJmdaaxngyjFYsP2ZJsN9rg7/DpF2W/36dnxLqFCxMXL5S9qM/PX8HIyBACgW6MjAzhlfkrOtgNhwd23CEB1Dy8RzxlH2bU3NvVmS617V7r7FAhXOjvC0JKua4mxHMjQ/B3+NDfX35krtvtcpr7ppNr1qOu1KFv7kPng6daV4LC4iCxKOhreTert+QP/FEMLd/O4oPBK5XngNag5d2s00C1wZ/X11PYci8+4rTlneWqNawqi7oe0egkpJTwd/jWzIoS7UZ7Mhgt3iYDyrfKLGsOmUy27AUhkUg6221F2dKB0CByuTxisSks2su4as2WBKuRyKgOBoHCSSgV5vupNy1gY0f+qS1pr9cDw2hDu9GKmeQl3Lj+atXrU7VkgUAXDKNVd98qqdRKfaVZyJKqN8Px6CS8Xg/C4dCmg+hK3IXzmdfKugYCXZifv1LytXMjQ7j73h3Mz19BODxQ8eecObKXcG5kiNtcVDdGYSSZ+hAnZR4DoUFIKcsaB9eqB6xmYuICACdoqbV2VAiXnqFcdn0XLxRmKcfWXI96/bBrnIlaCzUIX9VcHnh5DI/++ilnvFMho3j/qbZNBaLqdjYbiAIrR5yqYPrRL/bg0S/24MBrsZV5pEWX3f9avKT84GFsexmJ6SSEEPp5JtpL9mQwCpRuk60eWaKCyeJtNcBphvF3+PSLs2XNwbJm0d/vzBW8c/sN3Sjk9XpgWXMlc0ozmSzGo5MVZ5eOFz4Vr56jWQt1vUIIvDJ/BW6x0qAlhAsjVTroV2c6Mm+vdKf7O3yQUuKqNavnd6o3rGh0Em63C/Nzzpucuq/1xC0o2mlUwNkbPIPx6CSe6+yBZc0iHA6VZdzVB93VHe5rEcKFcKEp6bnOnk2f0BMIdGF+ztkpiMXKG4cU9zrrW9floAu4m8f+60k8eKoV9079Cu79nPM43m/vcgLJu/mVeZ4ffYgDfzBZ84inlneW6xOMfrv0vn/whSv44H/6d7j32aGycVUH/u2YM1WgcD/U7NW1eL0eBALdiMem+NpHe9KBRi+gUYz2Nh0UFo8sse1lWNZsxZmWUkq9vQY42UIhhA5aVXbTyTDmS7bT1Pa9yjyqoFV1dKsAeCOjl9Q228TFygXt1cbNqHWo+5fJZPWbnbrvC6k0pgvb/rHYFCxrDrnCYG8VlMe+NFX3Y0jV9WUyWb44044gCqUhkeFRRAsfDkdGhir+TQvhwo3rr647Mz9x8TxSqTRsewnPdfbglfkrm/r78PsfviWsaq3rOVlEZUT3vZUGfjaEj04GceA1Z/dl/3/4Xbx/4abevn/0S2f01n3LB+85DU3tXU6A+pBO9ZZvZ/HgU5t//XjwCee+FweeB758UTcpfRC6rJuZ9r2Zxr2eMSfD+9gTaLkr8dj438eD/QecIPUXz5cFyBs5hIBoN9mzwai/w6e3nqcTSZhmHDeuX9NB4epMxupxG6pDNhDo1i/SL5oxSCmxaC9joFA3OjIyhGh0EidOPlsIrNpg20voDZ51Rg0V3lic7ZnzG3pjmS6suVJdGFB9298wWp2az+ExHZSq+6c6cXWWOBwqbMuXb31vRX3T6k5gop0gHB5AODxQcuJSNRsNIl+Zv1LIjDoBaaWxZ/Wktvrr6qAL9z/jw7630vjY0NMAnO70B48L7F+cxf7FWWeb+24eD5704D58wPf+Cx58woMP+79Y87ikfe8s4YNffHgz0MPcf9qHBwcFDnxlEvveSqPl21k9Q3X/9aTT3FQIkD/6mQG0fOe2Dq4BoOX2LTz4jA8PPtW6ru56or1i7wajRQFUaiGNkZEh3aTj9Xr0jL7V1BuICk5Vp7ZlzSEanUQg0O2cjZ5KY2LiPE4HunUN6sjIEF4IhzBQaPwpnv83cfHChptnVJCrsizCLUqyvSrLuHq00wvhkL7P6r4Ur2Hi4gX0Bs/A6/U0ZFj+TLLySCeiZrfVweGN69cQGR4rvI4s78i/kw8+f1mfZPTgqVbnBKbFudIjRA+6Ks4erVU9mpfUOj74whU88vIo9r2VdmaQ9g85a34rvbIVX+jKB4B7PxvSc0YfPC62/Yx7op2k5e57dx40ehGNUjxG6cb1VzEQGoRtL2EmebksMIwMj8E0Y7j1+k14vR49EunW6zcBACdOnoLX68ErRRmEtbbgUqk0Mm9n4T3iQXt724YbaaTM49Dhp+Hv8OnsRW/wLGx7Sa9NfS0Q6Co7M1utBaic4ZRFtadERFtNN//skODtwMtjOPBaDN+bfHPHrJmo2ezZzCjgzOBTWUHTjMG2lxAIdFfNUBaPelJjkhLTSb1NHo9N1Ry01Wtru9rtFU8LALBmPdJaa2EQSkTbSQ2V3ynu/dL5HRU8EzWjPdtND5SOV0lMJ2EYbYiv0VVavBVW3Fmey+Ub2gUphHjoudFERLQ1HjpUn4jWtKe36QHo5h1VF7lWJlDKfMn3bXsZuZzc1DZ7PZw4eQqZTBZ3br8BAHrw/Y3r1xq2JiIiIqJa7OlteqC2Y9eU1QFns4wcajdaYdtLOljOZLLrOoeeiIiIqFH29Db9bnGkUD5w1ZqFlHmnu7699mHazSgxneQ58URERHsAg9FdQI2hymSyhSM85YaOG2wmkcgYeoNnG70MIiIi2mIMRncBNSDeNOP6fOOtGES/XWx7GW63C1JKPXSfiIiIdicGo7tEX18QUsrCWdgDO3okkxAujIwMbcmZ90RERNRc9nwD025xbmQI7sJs0WrHgu4UXq8H/d4gbHsZiQQzo0RERLvZnh/tRM3LsubQGzyD+bkrO7rsgIiIiKrjNj01LRWA2vZyg1dCREREW4XBKDUtIVwQQiC1wBFPREREuxWDUWpq7UYr540SERHtYgxGqakZ7W2QUnKrnoiIaJdiMEpNTQ3vZ3aUiIhod2IwSk1NNTGxbpSIiGh34mgnanoHHz8EIQTu3H6j0UshIiKiOmNmlJqe1+uBlBJS5hu9FCIiIqozBqPU9LxHPACATCbb4JUQERFRvTEYpaZntLcBAHI52eCVEBERUb0xGKWmJ4QLAE9iIiIi2o0YjFLT83qdbfqcZGaUiIhot2EwSk3PLQQAsIGJiIhoF2IwSjuGZM0oERHRrsNglJqeqhnNMTNKRES06zAYpabndotGL4GIiIi2CINRanoqM8pteiIiot2HwSg1PZUZ5TY9ERHR7sNglIiIiIgahsEo7RiSc0aJiIh2HQaj1FSkzGM8OtnoZRAREdE2YTBKTSWVSsOy5hq9DCIiItomDEap6dj2UsXTloTgiCciIqLdhsEoNRW/3wcAuGrN6q/lONKJiIho12IwSk1FCBe8Xg8WUmn9tUwmCwBoN1obtSwiIiLaIgxGqen4O3y4WlQ3atvLAADDaGvUkoiIiGiLHGj0AohWO+L1QEoJ215GJpOF+VIMAOD1ehq8MiIiIqo3BqPUdNoLGdBUYas+k8lCCKHrSQEgMZ0EAPT3Bbd/gURERFQ33KanpqOCztRCGrnCoPt2oxVCuGDby+gNnkUoNKhrSYmIiGjnYmaUmo5qYkql0nALFwCgw+9DYjqJSGQMUkqEwyGcGxlq8EqJiIhos5gZpaZkGG2QUuLDe/cghMD733u/JBCduHi+0UskIiKiOmAwSk3JKIxx+rs//qMY/NXPI/4vf4eBKBER0S7EbXpqSqqJKTGdRCaT5dY8ERHRLtVy9707Dxq9CKLVpMzj0OGn9b9VICoKNaRERES0O3CbnpqSEC6EwyEADESJiIh2M2ZGqalJmWcQSkREtIsxGCUiIiKihuE2PRERERE1DINRIiIiImoYBqNERERE1DAMRomIiKjppVJpSJlv9DJoC7CBiYiIiJresWeOI5fL487tNxq9FKozZkaJiIio6XmPeCClhGnGG70UqjMGo1RGvt+C3qtuHPznP4jEXxxs9HKoCvl+C1LvPIrUO482eilERFtuZuYy+vuCGI9ONnopVGcMRqnM1a9/DNbXHwMAhOZdyOT3N3hFpayvP4bOf/N96Pw334fxP/l4o5fTEPL9FpyYflI/DvzQQLS37MXaSSFc6PD7IKVEYjrZ6OVsyl58/tbCYHQTVGaqFvveSuPRXz+F/a/Vvr3Q8s7yRpe2KdPLHwMAiMeccmIVmDaD8T/5OHqvunVGMHr9CZh//vi2rqHlneWGPTfK+J98HJn8foR/7D0YP3APkT/+Ow1dDxFtr4HQIHqDZxu9jG13OtANAMhksg1eycbZ9jKzu6swGN0g+1sHdGaqlkDgkcQg9r2zhEdeHkXLuw//Izrw8hge+/Vn8djQ0ZouX0+rA2z7rw9s6+1Xk8nv14Fn/4/c1V/f7vUd+MokHv1iDx4bOop9i3PbetuKylb7PR/A6/qoIWsgosaRObmjA7KNEsIFr9ezo+tGU6k0/B0+/d+W1Zj3kWbSFMFoyzvLeCR2FgdeHgPubkPq+m5+XRnKSsb/5OPIve88fA/bIt23OFcSUO6zH/6Lt9+eBQC03JWbXutGhX/sPYjHHuj72Wipdx6FfL+l7Osqg7td9r2zhJa70nlubjRmq0gFoAPzoqky10SNJGUekeGxHR2orIdbuBq9hIbwdzhb9Tt1qzsnpf5v68tzWLSXGria5tD4KONuHo9+sQf7F2dx4LUYHnl5dFtu75GXR3HgDzaeJk+98yi8ro9g/MC9h172wGuxkn/ve6eGX7z3Vv7I9l/f3oBHZR3bP/kh3I/dh/ux+9t6+9UUr0N9ABCPPUD4x9/b3oUUPTc42Jg3A69wglEVnLf/wIcNWQdRMxkIDSKVSiMxndzSQEXKfFPULC7ajS0XapQjXg8AYHFxZwZxbxdltPdidruSuu9vZjJZvPnW12u+/JN/8xfw3V35lHD///0y/vDwP6j3skpvrxAMHvjKJK7t+xF8+Mj6aw4/0dIJ+1tuAMCz3/eX+MNrf1r1sv/Vwc/AI/4TXPKbAICv5z7Cm9f+eM3rb/2hn8IPf/0r+PCRJ/DIXYl/P/uVDa1zI37iQzcS6ETvVef+few7t3Dt2pvbcttruf/RIwB69L+f/b6/xD/85DLe+tPv4q1tXEf7D/woPG//ewC1PZdb4RiAZ7/v7+HV7/wwntj/Iboe/Q+4du2vt30dtTp58jg+/vEnGr0M2qFSqTSihRo7o70N/X1BGEZr2WUsaxYzycvw+30QW5g1TKXSiETG0N8X3LLbeBjhFjVfVso8XjRjWEilkZN5uIULwi1gGK04NzK0havcGu1GGwAgtZCG3+9r8Go2RmVHM5ls2e/yXlT3YNT/0z+Hv/7rb63rZ77zj5z//+bfAp/++Hdx+vQv13tZmvvRldv72reB7l/8Hzd+ZT9+Brj9Nbx6+2t4tYaL/+gnnNv/40svA3j5oZf9858HHvnwu/jat4GuzaxzIw7/KPDj/wPwnQziX72M+Pdy23v71fzw3wd+5OeB72Tw6lcv49UGrOsXjgC//6zz+xr8Zy/ja99e+7ncat/9mBv/c7M8P1X09v4D/Pbl32r0MmgHMs04IsMrO2aphTTsxSXMz18pvdxLcRhGGwKBri1fUyaTbXgQ5BYuvVW9VuBtmnHdLNNutMLr9UDmpA7eLWsO8djUjgqI1FrtOmWGx6OTDQvKZdGW/V5W/8zoN/8/XLliretn/tZ+CU9962v40U8A3zjsw+9Ob22H4H/+qvOH+R9/+lfwu//9ZrONP7Xunxio8XIv33sPT9x9F989+GQd1rlRrQAmGnTba2nsul6+55QGRH6+Uc/LztLTE2j0EmgHsqw5RIZHIYTAuZEh9Pc7mcjVwZeUeaRSaf39rZZrggBCZUbXyqylUmmYL8UwMXEepwPdZY+bCvSf6+zBK/NXdkxA6i7cd5mrz/NQr6C2VjmZh8w5ZSTcpndsSRvyut94un8a9/4ohgdPevBD7d3o2eo6vML6Pre1t0JEVBPTjCMcrvVj6t6h6jL7+4NrPj7OmeVSdyhvtbebKIDIrRGQtbe3wXvEU1JOkMlkdWY3HB6AYbSis6sHz3X24I1bN7e0vKFehHBBCFG3mtmJi+frcj21cgsX3njzrT3TaFeLxjcwAcBBF+59dggfnQw2rCGEiKgROHOwMinzsCxnqkj4+bUDdZVd2imZvXpQQWPm7eqBsRqDlEql9dcS08mS+aR+vw+x2JQzSD7R+KasWnm9nrptcXsLDVFbLZVKYzw6iT96bQG///tfxuOP13ZYifq54udxt2mOYJSIiKiI6pQ2jLaHBgupBedNeruCilwTjBRyC7VVvfZanECz9DKrgzg1SF49jjuBGmu1FVMT1IiwEydPIRQarNv1mS/F4RYCv/zLvwif7wR+4id+rOyymUy25D5Z1hx6g2dhmnF0dvXs2pmkDEaJiBool5NsYqhAZfxqmaVp20swCh3WVK64qUsHsUUBTy4nIYTYVDCfSqU3XXtp28voDZ6tqY6yuGa23hKJJCxrFvHYFMLh0KavL5PJwnvEU7i+ARz6wU/CtpfKSiykzKOzqwdHjx3XXzNNZzTkjevXnLU1wUixrcBglIiogVRQsFMHeG8VlfF7WICkaiC3ewB8vZpnNkq4nftbazOVChbV46kCoVQqjc4uZ2Re4HPVJxFImcehw0crbhWbZhy9wbN4rrNnU8FSJDKKVCpdU4CpHv+tqHG17SUI4Yy+qkfph2G0Ihwe0GsVbhdyudXZ6jwGQoOFzKiEbS8jk8nq8VXq9WG7sv/bjcEoEVETYFdtKRVkPWyepnqTNtq3LzO63kBU1QVHhsfq9qGjUoazGhVwJqaT+Jt33wUARKOTOHHylA5EJybO63FVielkxaBSSllSbwo4v7fj0UmEwwO4c/sNLKTSOHT4qP59Ho9O6q1ltV1d7Xc9tZBGu9HakLFZphnX6wwEumHbS2uudTNWP3dvZ7KIREZhWbM62LxqzZZMDRgolAtsV5PedmuOQ8eJiIgqeFjmSx2luN3NS7XWjWYyWTzX6QR8UkrInEQsNlW3ddQSGLe3txXOc185DTAxnYQQAhMXL6C/P6gf58R0EpHIGM6NDOHQ4aO4cf0avF4PhHChvy+oA1XVoR8ZHgMAvFDYzs7JvC47kTIP04zjdKALgUAXxqOTMM2Yvi23EGVTEmp9XOtdt2tZszDanTm1gUAXJi5eQGR4FKYZQyDQjfDzA3UNkqWUOHDACcFU4C+EwPzcFfQGzyIandRTG1Qtbzgc2pY5uo3AzCgRURNYa0TPbmFZc3UfZ6O281W2aTNSqTSOPXO8JCu42TrIaHQSUkrEY1Pwej0NOcJTCBduvX4TsdgULpwfAQD8ky/8Cu7cfqNk+xgAphNJ9PcHdbd6cWZQ1U8Wf822lyqeeKW6+KWU6PD7kJhO6mA4Gp1ENDqJyPAoeoNncejwUUiZR39fELa9VFOTjgp43es4iWotOZmEnHRKAAAgAElEQVRfdV8HcOv1m5i4eAG2vYTOrh6cOHmqLpltNZqqu/uUzoQaRhtemb8Cr9epLRVClGWn1WEFuxGDUSKiJrDba0Ytaw4DoUHnjb2zp+z7UuYbPr5m9basacZx4uSzOHT4aFlQWtx0Zllz+vlT2/HOZfK4as0hUOhWz2Sydc9srSdD2N8XxNDQCwCAx58oP7DDtpeRWkjDMFr1uLHiYO9qYdSW2ipWdY0qoEpMJ/U4ruKTnxYKx6caRhtGRoZgGG2YSV5GINCtDysQwoWJiQvwej2F35O1g3ZVc1mvmlF1MlUxd+HI1L5CFti2l+qybe/1etBeyOTfev0mblx/teTQAcNoxZ3bb2AmeRkzycs6C52YTqKzqwcHHz+E3uBZhEKDu2ZWKbfpiYgaSL2ZNsOpPlspMZ1Eu9GKWGwKphkvO8byRTOGaHQSphC4c/uNmq9X15auIyixrDkkppOYSV7SX7PtZd2V398X1LWQaot2IDSoO5oBlDSgRIZHMT93BUK4MD2dhPeIE5wtLi5BSgm3cOmTpF6oQ3f2ZgkhKg7uVyUPC6k07MJ/j0cn0d8XRGohDdOMIRwO6e1qFYSqADSRSOqpBur+er0eJKaTOuNnGK366M3VgbkQLswkL+PEyWfRGzyDW6/frLh+216GlLKuExQCgS5Y1hxSqTRSC2lY1pxuJgKcALyvP1iXchCv11NSqlHtOtXjEwh0IRwO6SNcVRY3A2DRXkYg0LXjG5sYjBIR0Zaz7SWdIVxdJ6hqC4UQZVvDD7PRjLI6l1294auMrMpYmS/FIaXUgVNxt35xU5Xu5ncL3QmuglE1nuqqNQcpJSYuXmiKE468Xk/FjKoqeUhMJxEOh2AYrQiFBmFZsyX1pYoQLsRiU4UMXQz+Dh9mZi5DCBdSqbTOLKou/mr3XR3napoxXcagHsNK1HNVzxrO/r4g/B0+mC/FkUqldfbSMNrg9/vqXpO83uBRdfbv1pPaGIwSEW2SacZhbLIL+GHDy2u+nlUZx2agArZq61K1hSqwWc9g743cV5VJigyP6mBUZfcCgW5ImUcikUSgcJ67UyvojNtRQYFlzcK2l0tu33zJ2TJdmZGquqYlwuFQUwUSlQJ+JzvoPA7nRoZ009JawWR/X1BnkosDrOK/hbUCuVQqjd7g2ZIMZCDQXRL0rmatKheoF6/Xs+1Hg5KDwSgR0SaNRyfh9XpKtnHrTdVUAtXP0lZjdCqtw2kgicNdqM1rxNGZ1ZqMVKOGv8MHIYSuW1zPda43QxoIdMM0Y0hMJ+Hv8OmOZcNoLdle7+zq0YGSajI7Heh2GnAio/iN3xgH4JQZqAxiJpNFKDSos3xCCHiPeEoCQLdbbOpDw2bKOsLhgYrBqNfrwfz8lbKv1/JcbHSb2O/36XIJlUldi2XNObM3O3y7trN8L2IDExHRJp0OdOns31qkzCNUGGy9Wq5oHE6lzOCLZswJnqqcH57JZGGa8Ypv0JHhMX27qvavEaoFULa9BH+HD16vB6cDXSVNTA/rkldB0HobS9T2eyQyhmghyA8EuuH1enTTTmI6iVwuj4mLFwAA04Wg2dkuDWHRXsZv/uaX4PV6EC18IHnj1k0EAt1ITCfhLmQWpZSIDI/i2DPH9f8OHX56U536tR4KUEl/X1Df/2bg9/tq3gpXH8hGmmj9tHnMjBIRbZIaXWNZc3or1raXsWgv6U5YwGnwSEwnddADlI+m6e09g9RCGhMXL+jrUjWVAKpuX4ZCg5BSltwe4GyDqi3nQKALodAgjmxzs4O6j5WaZlSjiOpYVo/ldI0n+ajtYNVwUyshXAiHQzo7CkA/dobRiomLF5CTEqcD3TCM1sJYppXbmLh4XmeoM5msrj91mnAuldyWE7guQebyEG4XvEc8NWUB16JKARqR4W4UKfNwF+pUGzEYn7ZOy9337jxo9CKIiHYyKfP6PGnVCd4bPAvLmkUsNqWDnGPPHEcmk8Wd22/qLVrbXsaJk886NYXPD+DYM871+Dt8esvUNOO6O/mNWzd1g0hv8CzC4QG0G23oDZ5BINBdFgiFQoNITCdx4/qr+rhBdR3bxTlK8mmICp3y6nG69frNwmxL57JK8eNXzbFnjiOXy6+rC1+JDI8hUZituZPqBVWW+3ShrpVoJ+M2PRHRJgnhwulAF6SUOsumMnWRyFjh385MRsNoqxg8yJws2Z5PLaR1HaS6romJ8/pn1TD1aOFUGwAVgzan/rINphmDbS8VBmpvb/AihKvQGLTqPhZG1RRnilXTjFLLMPuRkaHCVvjYutc2cfE87tx+Y0cFooCzPd/fF2QgSrsCg1EiojpQY4sWCvWOqoZRSgnTjOst3vZV26oqCMvJvG6iUcHYVWtWD053ftaZq6iaOEQhUCtuvlktk8nCtpdw1ZrDxMULDWv6UDWKvcEzGI9OYjw6qc85Dz9f2mXuNFi1wd/hq2k7Vo/lMWO7Zgg40V7CYJSIqA7UuCBVO+p8zQlQx6OTOjitVq9p20tIpdL6lBrA2Z5/0Yzpbu5Fe6mkq34meankOMHVzSzFHebx2FRDRwsZRqu+X+o4SAAV6/+EcOHG9WuYn79Sc+Zvfv4K+vuCiAyPFupnd/eJVkS7CYNRIqI6UcGn2qr3ej0Ih0M6OwqsZDcVdU61Ou1FBbWG0QbbXkI0OqlPmgmFBvFcZw9se0mfhKNus1IGUQiXDlBTC+nCPM88bHsZljWHUGgQJ06eQig0uDUPyCrnRob0GekTFy/glUIAWS+x2BTm55w62916hjfRbsQGJiKiOslksroBCQBmkpfh9/tKGnJUo06xzs4evdU+P3cFfr8PUuYRiYwik8liYuICMpkseoNnAKDkpBuVKQ0/P1BxzE8qlUZnV/lZ8IAz/7LdaG3Y3FEiIoDBKBFRXanucGAl8LSsOfQGz8Aw2ioOpFffL+6gr2YjJyyp0UOWNYtFexntRqs+5YYNMETUaAxGiYjqSGU0j3g9JYPF1Uk/1YaUP+z8biKi3YrBKBERERE1DBuYiIiIiKhhGIwSERERUcMwGCUiIiKihmEwSkREREQNw2CUiIiIiBqGwSgRERERNQyDUSIiIiJqGAajREREe5yUedj2cqOXQXsUg1EiIqI9LJPJ4rnOHpw4+SxOnDzV6OXQHsRglIiIaI+SMo/e4FlkMlncev0mAoEujEcnG70s2kK2vQwp841eRgkGo0RERHtUIpGEbS9hYuI8vF4Pzo0M4XSgu9HLanr1DOakzCOTya7rZ2x7GaYZ39DtmWYMV63ZDf3sVmEwSkREtAdJmcd4dBJer6ckADWM1gauqnmlUmmEQoM4cfIUDh1+GgcfP4RDh4+iN3gWienkhq/3RTOG6Dqz0ZvJXueaLCsKMBglIiLak65as5BSoq8vCCFcVS+ngtbd3OD0sOykacZhvhRHX18Q8dgU7tx+E7dev4l4bApu4UIoNLjhgNS2l3HE61nfenMSwl39OXvYzwLOfbKsuQ1dR70xGCUiItpBQqHBugQR6jr6+4JVLyNlHgOhQUxPJ/FcZ8+mb7NZRSKjOHHyVNWMY2ohjZnkJfj9PhhGK3I5CdteRiDQhVhsCuFwCKHQIDKZLDKZLCLDYzXftszJmrb9i7flM2+vb1u/knB4oGmy4AxGiYiIVpEyD9OMb2r7daukFtKbDiKkzCOVSsMw2uBdIyv3ohmDZc1i4uIFhMMDTfl41MPIyBBuXL+GaHSy4n10C1dJwJhaSMM0Y/rf50aGAADmS0620TRjNdeV1rptbttLOjudy+Uhcxvbbi++vbWe++10oNELICIiajaRyKgOSixrDjPJSw1ekUNl3jZrcXEJUkp4vb6ql1EBeSDQjUCgC36/Dy8WBWCNJGV+zdKC9VJB2cjIECKRMfg7fCWBWiw2VfYzxUGdEC709wWRSCRx4/o1CLer5vVJKWu6XPEa3Bvcole3t9FAdqswM0pERLRKaiENIUTFIKSR6tXFrTJsa2XGEokkpJR6G18IV9PUGCYSyS1ZywvhENxuV00NRas/FHT4fZBSIpPJrln6UC+5GoPYSu7e/V7V36VUKr3tzzODUSIioiK2vYxMJot2oxX9fcGmyYoCQC638QCkmG0vAQD8HdUzo5Y1CyEE/P6Vy7wyf6Uut79Zfr+vZF31IoQLgUA3rj4kGHMLUfa1tbKNUuZhWXNIpdKVv7/O59V7xKM/UNQ6N9Sy5tDZ2YO/+qvbGPtfonoiQCg0qNfY2dmD3uBZ9AbP4NxoFMeeOV5yH3qDZ3HsmeN1b2ZjMEpEtEWabbA0VWeacZ3pUgGD0d7WyCVVpBpXNvu7pbaYq9We2vYyUgtp+P2+ku3mem6Nb4ZhtG7ZWvwdToazWuAIOI+DlKWNRyrAr5RtftGMYSA0WBbEpVJpfOc7OfzZn/05xqOT+ndQyjwiw2PoDZ6t+Fwb7W1IpdIwzTie6+zBocNPIzI8BtteRm/wbMU1e70ejIwM4cknP4F/8oVfwY3rr2ImeRmJ6SRMM45IZBR9/UG8cesm+vuCmJ5OIpPJQso8pMzjuc4epFJpZDJZ9AbPPPyBXAcGo0REdaDG3zgzCI8WZhA6mYdqcwjVz/QGz254gPV22OlBtcrodFbpBrftZUSGR9HZ5XxfbX82S6dxse2q9VOB2Hq3m217GSdOnlozkKtGBWAnTp7S3egquKr1+qTMIzGdxLFnjuu/vfVSGdfUQvXbVAFncfB41Zqr2hA2PZ1Ef38Q4fCA/pppxhGNTmLfvn24/+A+bHsZx545Dsuaw3OdPTDNGGx7qeIUAxUMj0cnEQ4PIBDohmnG8Nu/869hWbP68bLtZR0AG0Yr/H4fvve99/H4E4/DMFoRCHShvy/o/MxCGu1GG4RwIZPJ4vu//0kAQG/vGZw4+SwymSzisSlMXLyATCaLEydP1a2hjcEoEdEmqbO9VZ2Z3+9Df18Q4XAIgUA3LGsWodBgWTD0ohmDacaRSqURGR596LngKhMSGR5DKDSIyPDYlnc3J6aTOHT46R3dRZ1KpZFKpTEyMlTxfiwWMlqqOejtQoBRaSt2t3jYtrDK8lUKyFVjU6WvP9fZA7dwITI8hsjwWMnlVm8nr84SDoQGkUgk4RYu2ItLTnAaGUUmk10zMCwWiYwiFBqEv8OHmeRlWNbsmlvKlT5oCeGC1+tZ8+fcbud3Q5VNjEcnS+prE9PJkuv2d/hgmrGSkU+G0YqZmcs4csSDp5/+DGaSl9DfF0Rv8Axsewn9fUF9ROvq1w5VXjExcR7nRoYQL9Q2Ly+/DmDlsR4IDeLEyWfX/ECZk3nkCnNWM5ms8/eykMYv/MJnMTIyhEV7GUIIzCQvIRDoQjg8gImLFyClRCg0uOGgvxiDUSKiTVI1Vzeuv4ob169hJnkJsdgUzo0MYSZ5CXduv4lAoBuphXRJMPR2IdNw5/YbGBkZgm0vVc3eAUDPL/bhwvhF/Wahglx1+1thoZBhiURqn5vYbBLTSbQXskKVMn3F2cZmadDZamqbXgVVqy0WAjH1fTU7s7OzB23/7U8iMjxalq180YxBSolYbArx2BQSiaQOIlWgqrrxLWtOZ9vU9VvWLPr7g5ifv4L5+Ss6CD03MqRHJ61FfTibuHgBsdhUyXUrphkvCcxOnHy24t+P94hnzakFqkQg83bWGYhvxmAYbejvd36/otFJRCKj+vJqFqlpxnDw8UOIDI/pEgh3IROpLmcYTnlIoHAq1rmRIWTeLp2i4Pc73f7qA/DqtaYW0oWyAOdDxVrHf9r2EgKBLni9HkSGRxH6/CCEEOjvC+LcyBDu3H4DN65fK6nRDYcHcOv1m7j1+k3MJC8jEOiqev21YDBKRLRJqYU0Ji6eL8kiWdYcDh1+GplMFkK4EI9Nlbx5AIBwCx2cnhsZqhiwFjt48CD+8JV/hxvXr+k3ghvXX8VijQ0MG6GClmZpXNkI215as/4zJyWEEBBCILWQ1ve5kfWRKvu4OsjYTAf1emQyWRiFLVu1HntxCU9+/yfwN3/zrv5asenppN6mdmoNpd6yVqOk2guB1ur7VSkoXisYXB0Iq9me4XBIb4WrQFgFUaocI5FI6uvIZLIPbVYqvs3iTK/L5ULid34PkeFRGEYbZpKX9ONVqQlq4uJ53Hrdqcc0zVjJB8/i+6oeM/V6ImUeuVy+LDs8MjKkt8tV4PsTP/HjEELAsmZhWbM6oA2FBvVA/3B4QGdWpcxjZGQIL4RDmElehhAC3iMe/Xr1MF6vR2/1bwaDUSKiTYrFphAZHit5g7SsWRhGm36TVW9SuVzpbMLin1FbbdXqR5944vGybmrDaHXmGjZJY0kzethcTttehtvtQrvRilQqrbewK70ZqwCmVsXbtap29dDhoyXNUquvT22vJqaT6OzqqZit3WxXvZptWe33ZvXsS7Wl/PWvf3NlDUWXUVu8fr+vEEg7GVDvkUJtZaHxSt2eCqzUY6zmdJpmTH8YCwS6YBht6A2e0acbKdHoZEkwrH4m/PyA/re1KhuosoPqGE3zpbi+r6ufg8zbWbiLHhvLmkNnVw+E26UPDPjwww+w8P/8CfwdPrwyf6Xk9yX8/AD8fp/e6lfjktxugY5CVlNln1cPvVe/f+r+pVJpSCnLykb6+4J6RyW1kIa/w4fI0AvOSVEdPowUtu9vXH8VExcv4HRRplUF6OpxF8KlX0vm569sOtO5XgxGiYg2qb9wXvXqTJHKiqr/zmSyJW9YbiFK3vTVG4NtL1UMeHK5/LafD67WW6+RQsBKk8vBxw81RS2qbS9BCLEyJ7LKUYsqODz2zHEce+Z42ZGPodBg2e+Aacb11vR4dBJW4Tx4216GacbR2dVTtk2cSCSRSqlsuxOM1ftxEoXAplpGXQhR8juoAmTbXtLbyMXUZb1HPBiPTupgUwWs04mV9dv2sv4QVnz7sdgU/B0+hEKD+gPZK/NXdK3vsWeO68cqt2rovVqXqvUsLitRR3eq65S5fOFc9ln9+138XKq/1UrZ9Gh0Es919qCzqwcffPAhAoEuzMxcLgvqvV4PZpKXdHbT+vIceoNncOjw0wiFBpHL5TExcR5AIVNZtAUeCHSXDN0XwgUhRMX63XMjQ7orXq3D7/dhfv4Kzo0M6SCzmY7+rITBKBFRHajOVEW4SwNN9WZXvJ1VKejpKLwpVcqGffKTP4B/9a8ShaBmri4n8ahtPtOMVwx01ZtsPYPgSGQUtr2EcDi0LbWoqwOrYrLQuOH3+8q2kFdvHVuruqVNM4ZDh4/q71cLGN1CFLZ4YzoIXLSX9LbpYtFjqyYs+P0+TBdl99Tjr9a42efDXfQhqRJ/ITAPFTK0z3X2wLJm4e/w6W3w4lpbFVQ6Y4Jient4oZARVMGpM67ojP7bWFxc0vfHtpcRi02hvy+IyPCofnxOB7px9707zglH00n83u9dKclaKplMFuOFYNHtdmF+ziktOfbM8ZKOdPOlGMajk/B6PZifczKaljWrg1X1PBbPYF29FT0yMoTf+I0o4rHfrGlXIvz8QKGEIIRYbEqPTwJQNss2HB7A/PwVfb1+vw93br9Rddtcvfbs5N0RBqNERFtAvTEsLf0FIsNjZW/kQOWO5rU6uI8+/V/jrf/4DUSjk+gNntEZuo3Wi5pmHCdOnoJtLxW6+Z/VgamyEvwsbeg2Kt1maiGNcDiEiYvn4fV6ymrx6s3JllVev8pa+jvKh6ivfnO37SW4hQu3Xr+J+bkriMWm9DxKFdRVCghsewnR6CSEEPr5dxppnFrV4g8taks2k8kiMZ3UQZ2qc1Rr3GztqMr6Vct4T1x0snaJ6SRChYxof18QMzOX9e+oKDqS0u/3wTDaYNtL8Hf4EC8ElamFNHqDZ+Dv8OmRQLlcHrHYFIQQ6OzqwRenfqvwu/csjj1zXAeDphlHPp8v+boQAj/5k84g9uLfe9XdHY1Oot1oxfzcFfj9Tke9YbSh3WjV3erqQAMViMa+5JTHqABYfehYvVUdi00V6rSv4dzIED4fOlNzAOj1ejBx8TwmLp7X2+K0gmfTExHViVOnVnqO+eAX/inS6T+FEAIjqzqCVa1Y8fa9ri2tEGx8+oe9uHfvHu7cfrMwZ3DjjTZqnBSwsi1o20tITCcRGV7SjQkq+LlqzWGzp5JLmYf5kpMdVN3RXq+nMNhbQrhdW3KMYnG38ur1WNacPmVIjfRZXU6h5HJ5GIbQ65bWSi1otZ8BnMdOSolwOAR/hw9ROHWKgUA3vF4PTDOmf14972rbOR6bwkDIqUFWl3FmSsZr6jCvpnj7uNJJRl6vB3duv1nokM+jvy+of0YFnqoeFHB+B29cv1ZyHaozPFcYeeT1ehAOD+hz5Z0a0ThyUuLue3dgWXN6zJa78Jx4PE9hfu4KrC87OwXh5wecjOaqhrpAoAt3br+pH6OVRqKukqDS73dqKd1uUZJ5vHH9VVy1ZvFX7/wn2PZS0x0Du9sxGCUiqiPLmkVkeAw3bvwZACCd/tNC/djlqjVbxYHMWtuvalbhVWu2rtmVcyNDem3hcAgnTj5byMp1lQRoclWd3nqpDGI4HNLB9OqmlK0IRo32Nj2loPj6E4kkbHsJI4XaOsB5jBOZpN5OL+Z2u/R9cBcmIahA9mrh6MxKVOazvy+oM5FCCMRjU3r73rLmEA4P4HSgG3Z4GUK48ELhcYrHpjAeXQmSJy6e1xnxjT4fpwPdCGEQphkrCTSLCeGqGPBWCjyrKd4JKP75St9fHTgqtR79qeojH6bShwbDaIVhtKKzswder0c3+9D24DY9EVGdqBoz04zh5s2vAgA6/rufxPzclTXfJIsDUFVb116hSUS9IVca+bPutfp9egu4ODjQjShFb9jqfqn6vo0qrsUrbogJhwf09ulWNDSpgKq4PtW2l3Xd4AvhkP66qtmtFPyoreBjzxzXY7visSkI4UJ7YSt4NbdYGfWjTsAZKcyfdc5B70I4HCrpNp+4eF43nxR/Ta3J6/Xg1us3N/XBQIiVmsoTJ58ta8bai0wzjkV7GbEvTXEbfZsxM0pEVCcqSEgtpOEWAv944Ffh/+mfqrp9qzvVC5kzy5rTHb6VgiHVVau67WuZA7iW4mMA+73BkpE8xc0bHX4fEtNJfVb5ZjnZ41GdJT03MqSzpOr0nM3et2JCuPTA8RMnT8Hv9yGRSEJKqYNJRWVOi++/Egh0YSZ5GakFJztanMkzjNaKW7sjI0MQL8V1DSaAsmxj8fdqVY/Hp3h72raX6/I7tVOpxrHisUe0fVruvnfnQaMXQUS0G6ktv2r1Z+PRSUSjk86xoZ/r0rWTqqO4EtXdXI9TT6TM49DhpwE4wU0ul9e1jMVBmrqcEAITE+c3vJVuWc54G0UFosXBoGnGdQax3pzjKZ1ge7P3hXYXKfN40Yxtqg6XNo7BKBHRFokMjyGTyZY0NBVTwWgx1WW+Fsuaq9tQamcm4yhSC2nd7a1qFRUp8zh67DiklBip8WjGatRIKqcJpjFzD217uaTJhYgai8EoEdEWKT6VphLbXkZv8AxyuTzajVYEAt0VGz4aKTGdRCQypptw1sraEhFtBGtGiYi2iNfrWbMGzzBacev1m9u4ovUxzTgiw6MQQuiB4/UYtE9EVIzd9EREVEY1dAghMJO8hHCh43y7jyMlot2PmVEiIiqzuLgEKaU+mUgN2FdnihMR1Qszo0REVKZdHRdZdMqT1+uBlHLDx48SEVXCYJSIiKpijSgRbTUGo0REVKY4E1ockAohOBKJiOqKwSgREVXUVxjh1Bs8q2em8nQaIqo3zhklIqKqTpw8Bdt2zqQXQuDG9Wt79shIItoaDEaJiGhNphlHTkr09wUZiBJR3TEYJSKiPWU8Oonp6SS8Rzzo6w/yRCmiBmPNKBER7RmmGUe0MMw/tZBGKDQIy5pr9LKI9jQGo0REtCekUmmMRycRDodw4/o13Ln9JgAnU0pEjcNglIiI9gR1lOm5kSEAzvgqf4cPtr3EQf5EDcRglIioyUiZrxoc8Wz4jUstpOH3+0rmpKqGLA73J2ocBqNERE0ml5M4dPjpsgBJyjwikdEGrao+pMxjPDqJ3uDZqtvjUua3JOjOZLIwjFZEhseQmE5CyjxSC2kAaPgg/1Qqjd7gWSSmkw1dB1EjMBglImoyXq8HgUA3oquCtWbJ3qVS6Q2vJZFIIhqdhGXNIhqdxKHDR3XgqQLV5zp78FxnT80BaSaT1f9bi7sQcKZSaUQiYxgIDSKTySIQ6F5zZJVtL29pRjqTyaI3eBa2vYTpBINR2nsYjBIRNSHDaMVVa65kuz6Xkw1ZS2I6id7gWQBOwNgbPIsTJ09t6Lpy0rkPzrGiAlJKPNfZg1BoEEePHUc0OolMJot4bAqG0frQ6+vs7MGxZ47r/xUHt6sZ7W2FwHIJUkpY1iz8HT7EY1Nr3oZpxsoy0lLmEQoN4tDhozh0+OiagXAmk8WJk6eqXsbJ0kqMjAxhfv7KQ+4x0e7DYJSIqMHGo5Nl44VOB7ohpcTi4pL+WubtxmRGVSYzk8licdEJ5Daqvy+og1B1PVJKHZD5O3y4cf0aAoGuh16XlHlk3s4iFpvCTPIyZpKXcTrQBdOMVby8v8OHVCqNQKAbADAyMoSZmcsQwlVSPhAKDZYEtDmZx2LRv6XMYyA0iMR0EuHwALxeDyLDYyXfL/4QYb4Uh20vwbaXEQoN4tgzx0sCU3XZ04V1Ee01DEaJiJqAql1U1LZxcVDkFgK5be76tqw5ZDJZCCHg9Xp0QNxeQ9ayEttehiyc5uTv8MEw2gAAhtGGG9dfxfz8lZpPeRLChVuv30R/XxCBQBcCgS7EYlOIVcl0+v0+SCn18ab+jpVmpvHoJKLRSXi9HhhGW8kaVCZVSSSSsKxZjIwM4YVwCJlMFpY1q7/f23sGR48dB+BkRROFrffx6CQS00lkMtmSEgzvEee2XqwSRBPtdgcavQtv714AAAYgSURBVAAior3udKAbV4uCGWCloUYFTra9jMu//X9tKiu5EYuF2w+HBwAAMucEw0Z724auT2UEw+EQDKMVUuZx6PDTcAtXTdvySiqVRmohjfZCMJuT0tmCX1yC0d6GiYvny35GCBf6+4K6Sci2l+H3+2DbyzDNGPr7gpA5iYgZK9TtOtnZXK70A4D5kvP9F8IhJBJJ/ZxkMlkkppP6g4UKOtX3bXsJhtEGv9+HRCIJFXqGwwNITCd1MMwToWivYWaUiKjBDKNVz74sJgqZ0FQqjec6e/DpTx8pC4y22kLKCaxUgKRqPlU2b71Wd6+nCtff4fet+3qi0Un0Bs9gIDSoM40jI0MVH0slHA6VrUV9EMjJvA5U1TZ8JpOFlLIkU5rJZOHvcALK8eikfmxOnDyFaNG/Q4XrUGUBQgjEY1M4NzLkZJmLtupvXL8Gf4cPodCgfkyKJaaTCIUGOQ+VdiVmRomImpTX68FfLN9Cb/CsDmD+xb/4lyWXSUwnYdvLFTOBm2WacaQW0vB6PfB6PZAyj+lCsJYrytCqUoL1ZDbdblG4DSc/qDKctTpXCDozmSzcblHzaCbDaIVhtDlb74WGsLcLQaFlzcIw2nBuZAi9wTMIhQaLfs5ZnwoGr1pzuGrN4XSgCxMTFxAOh3DVmoVbCITDA0gtpPVjF49NQYhLkDKv13nj+rWytc3PX0FnZw96g2dx5/YbJd/zHvFgAcDi4hL86wzciZodg1Eioi1w7JnjmLh4oaZGnEpSqTTeffdd/NVf3UY4HHICnNTK9q/KrEUiY5BSIvC5rroGKZlMVs8BFULoNalsnmnGdQZyIDQIv99XU0CsMowvmjG4C+fDBwLdG36caq0vLRYODyAUGtRNSX19QSdwPOJBLDYFr9eDu+/dgWk6jUc5mdf3TQgXZpKXkZhOwt/hQ39/EKJQYlAcjN96/aa+jApAHxYwqwkFUkqkUumS59Pv9zEIpV2r5e57dx40ehFERLvNocNHMTFxvmL9XyqVRnt7W9XgRMo8jh47jvff/x6AFnzn29/UXz90+GnMz12B3+9s6Trjn5yGoGqNOxvRGzyrg1+322kU6uzsQWohrbvhJy5eAABEhkfLtsdNMw7hdpXd/0wmi2PPHNf/FkLglfkr68qq1oMK/CplKBtlPDqJ6ekkvEc8usufaC9gMEpEtAU6O3vg9Xp0gBgZHtNnoXd29dQUPP7q4D9FPH4Zd26/qQOTQ4eP4tzIEAyjVV9PaiGNXC5ftrW7UaYZR2R4FBMXL8CyZrFoL6O/PwjTjCEQ6Eb4+QF0dvXoywshMJO8pDN3Kpju7w9WzJaqhiHAqfHcSHazHmx7eduDYCIqx216IqIt0OH3wTTjumPasmaRy+XhLmx5G0U1klLmkUgkSzq4AeCTn/z+suv1ej2wrFlY1iyEEIWu9DZEhkeRmE7WpRNbNeWoGZq9wTMwCx3mExfPw+v14Mb1V9EbPINcLo9zI0MlW8hqRFG1tRhGa12zuBvFQJSoOTAYJSLaAu1Gmx7m3m606VpLNY+yOBuo5lauDtBUU486Ux1YqXcEnKyiqlVMTCdhWXN1CUaLg8tAoAs3rr+KRXsJ/g6fXrdhtOLW6zdLmnIA6CancHiAwR4R1YTBKBHRFlDBnBoar6hxQsWBWubtLIz2trLtar/fCf5SqbS+vAo22422kuuYSV4qG5y/UWqmqLK6OafY6rpGNSbphaIRSkREa2HNKBHRFlFzJoUQcLtdOih1ThtaaZxxTvCZKwsCAWd0k1uIDXebbzfTjJeVGxARrYXBKBHRFlHd74CzpX460A3TjDW0aYeIqNkwGCUi2kK2vYxcTnJGJBFRFQxGiYiIiKhheDY9ERERETUMg1EiIiIiahgGo0RERETUMAxGiYiIiKhhGIwSERERUcMwGCUiIiKihmEwSkREREQNw2CUiIiIiBqGwSgRERERNQyDUSIiIiJqGAajRERERNQwDEaJiIiIqGEYjBIRERFRwzAYJSIiIqKGYTBKRERERA3DYJSIiIiIGobBKBERERE1DINRIiIiImoYBqNERERE1DAMRomIiIioYRiMEhEREVHD/P8C4Br/CHB3CQAAAABJRU5ErkJggg=="}}},{"cell_type":"markdown","source":"# Imagine having these random datapoints generated by 3 Normal distributions A, B, and C","metadata":{}},{"cell_type":"code","source":"data = np.array([[10.0, 5.0], [2.0, 1], [3, 7]])\ndata = pd.DataFrame(data, columns = ['x', 'y'])\nsns.set(rc={'figure.figsize':(10,10)})\nplt.scatter(data.x, data.y, s = 150)\nplt.show()","metadata":{"execution":{"iopub.status.busy":"2022-07-17T12:41:21.849239Z","iopub.execute_input":"2022-07-17T12:41:21.850110Z","iopub.status.idle":"2022-07-17T12:41:22.242778Z","shell.execute_reply.started":"2022-07-17T12:41:21.850052Z","shell.execute_reply":"2022-07-17T12:41:22.241222Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"## Initialize some centroids at random","metadata":{}},{"cell_type":"code","source":"mean_a = np.array([8.5, 5.0])\nmean_b = np.array([2, 2])\nmean_c = np.array([1.5, 5.0])\ncentroids = [mean_a, mean_b, mean_c]\ncentroids = pd.DataFrame(centroids, columns = ['x', 'y'])\nplt.scatter(data.x, data.y, s = 150, c = 'red')\nplt.scatter(centroids.x, centroids.y, s = 150, c = 'black')\nplt.show()","metadata":{"execution":{"iopub.status.busy":"2022-07-17T12:41:22.245566Z","iopub.execute_input":"2022-07-17T12:41:22.246838Z","iopub.status.idle":"2022-07-17T12:41:22.506393Z","shell.execute_reply.started":"2022-07-17T12:41:22.246783Z","shell.execute_reply":"2022-07-17T12:41:22.504867Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"### 📏 next we compute the euclidean distances of all datapoints from the centroids just like kmeans 📏","metadata":{}},{"cell_type":"code","source":"clusters = ['A', 'B', 'C']","metadata":{"execution":{"iopub.status.busy":"2022-07-17T12:41:22.509124Z","iopub.execute_input":"2022-07-17T12:41:22.509560Z","iopub.status.idle":"2022-07-17T12:41:22.517382Z","shell.execute_reply.started":"2022-07-17T12:41:22.509521Z","shell.execute_reply":"2022-07-17T12:41:22.515699Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"data['A'] = np.sqrt((data.x - centroids.iloc[0].x)**2 + (data.y - centroids.iloc[0].y)**2)\ndata['B'] = np.sqrt((data.x - centroids.iloc[1].x)**2 + (data.y - centroids.iloc[1].y)**2)\ndata['C'] = np.sqrt((data.x - centroids.iloc[2].x)**2 + (data.y - centroids.iloc[2].y)**2)","metadata":{"execution":{"iopub.status.busy":"2022-07-17T12:41:22.519567Z","iopub.execute_input":"2022-07-17T12:41:22.521115Z","iopub.status.idle":"2022-07-17T12:41:22.543914Z","shell.execute_reply.started":"2022-07-17T12:41:22.521065Z","shell.execute_reply":"2022-07-17T12:41:22.542939Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"### Now we try hard assignments\nIn hard assignments each datapoint gets assigned exactly one centroid, even if it likes several centroids, we are being \"hard\" and make it decide!\n\"Note, likes isn't very rigorous, it's just meant as an analogy to the probabilistic setting where a point could have non-zero probability of coming from several clusters\".","metadata":{}},{"cell_type":"code","source":"#idxmin is a neat pandas function that gives you the min index along an axis, idxmax is its counterpart, wasn't in my pd np cheatsheet\ndata['Cluster'] = data[['A', 'B', 'C']].idxmin(axis=1)\ndata","metadata":{"execution":{"iopub.status.busy":"2022-07-17T12:41:22.546086Z","iopub.execute_input":"2022-07-17T12:41:22.546972Z","iopub.status.idle":"2022-07-17T12:41:22.589722Z","shell.execute_reply.started":"2022-07-17T12:41:22.546927Z","shell.execute_reply":"2022-07-17T12:41:22.588423Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"## Hard Assignments are boring, lets let our datapoints be a bit promiscous 😘","metadata":{}},{"cell_type":"markdown","source":"resp here stands for responsibilities, they are the soft assignments and also where the \"expectation\" in expectation maximization comes from. \n\nYou can think of resp_ij as E[Zij] the fraction of times we'd see x_i from cluster j from all occurances of x_i.\n\nI.e. what is the expected occurence of x_i from cluster j? \nWell simple! p(x_i | cluster = j) / (total p of x_i)\n\nAs a motivating example of why you'd want soft-assignments, imagine we're clustering students heights at ETH Zurich and we define two clusters, foreign students and students from switzerland.\nMaybe the mean height of students from switzerland is 1.85 and that of foreign students is 1.7. So if you'd see a student with height 1.95 you'd think they're more likely to come from switzerland, but you can not completely rule out that they're foreign either.","metadata":{}},{"cell_type":"code","source":"# we create a new dataframe so as not to change the above to compare later\nresp = data.copy()","metadata":{"execution":{"iopub.status.busy":"2022-07-17T12:41:22.591197Z","iopub.execute_input":"2022-07-17T12:41:22.591607Z","iopub.status.idle":"2022-07-17T12:41:22.597746Z","shell.execute_reply.started":"2022-07-17T12:41:22.591565Z","shell.execute_reply":"2022-07-17T12:41:22.596617Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"### Initialize Clusters at random","metadata":{}},{"cell_type":"code","source":"#initialize covariance matrix with variance 3 (random) and covariance 0, doesn't have to be this way it just simulates randomness and makes it easier to illustrate\n#when we initialize the off-diagonals as 0 we basically say that our clusters are perfect circles with radius rand_init\nrand_init = 3.0\ncov_init = np.array([[rand_init, 0.0],\n                     [0.0, rand_init]])\n\n#these are the normal distributions we initialize with mean defined as their center and covariance matrix above\ncluster_a = multivariate_normal(mean = centroids.iloc[0], cov = cov_init)\ncluster_b = multivariate_normal(mean = centroids.iloc[1], cov = cov_init)\ncluster_c = multivariate_normal(mean = centroids.iloc[2], cov = cov_init)","metadata":{"execution":{"iopub.status.busy":"2022-07-17T12:41:22.599249Z","iopub.execute_input":"2022-07-17T12:41:22.599644Z","iopub.status.idle":"2022-07-17T12:41:22.636434Z","shell.execute_reply.started":"2022-07-17T12:41:22.599608Z","shell.execute_reply":"2022-07-17T12:41:22.635210Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"## Play around with the cluster parameters to see how they shape up get a bit of a feel of the covariance matrix and its meanings","metadata":{}},{"cell_type":"code","source":"samples = cluster_a.rvs(size = 100000)\nsamples_b =cluster_b.rvs(size = 100000)\nsamples_c =cluster_c.rvs(size = 100000)\nplt.scatter(samples[:, 0], samples[:, 1], alpha = 0.2)\nplt.scatter(samples_b[:, 0], samples_b[:, 1], alpha = 0.2)\nplt.scatter(samples_c[:, 0], samples_c[:, 1], alpha = 0.2)\nplt.scatter(data.x, data.y)\nsns.set(rc={'figure.figsize':(10,10)})\nplt.show()","metadata":{"execution":{"iopub.status.busy":"2022-07-17T12:41:22.638003Z","iopub.execute_input":"2022-07-17T12:41:22.638432Z","iopub.status.idle":"2022-07-17T12:41:23.691803Z","shell.execute_reply.started":"2022-07-17T12:41:22.638389Z","shell.execute_reply":"2022-07-17T12:41:23.690526Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"### Now the responsibility A can take for some point x_i is p(z_i = k | parameters of cluster k, observation of x_i) = p(any point being from cluster k)P(x_i | cluster k)\n### this translate to prior * likelihood\nSince we have obviously no clue in the beginning we just use the most naive prior there is, the uniform prior (which is often a good starting point), i.e. p(any point being from cluster k) = (1 / #clusters) which you can also think of all clusters having equal \"weight\".\nIn our case P(x_i | cluster k) is a pdf, it doesn't have to be though.","metadata":{}},{"cell_type":"code","source":"uniform_prior = 1 / 3.0\n#likelihood * prior \nresp['A'] = cluster_a.pdf(data[['x','y']])*uniform_prior\nresp['B'] = cluster_b.pdf(data[['x','y']])*uniform_prior\nresp['C'] = cluster_c.pdf(data[['x','y']])*uniform_prior","metadata":{"execution":{"iopub.status.busy":"2022-07-17T12:41:23.699680Z","iopub.execute_input":"2022-07-17T12:41:23.700049Z","iopub.status.idle":"2022-07-17T12:41:23.711419Z","shell.execute_reply.started":"2022-07-17T12:41:23.700018Z","shell.execute_reply":"2022-07-17T12:41:23.709860Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"You might notice that bayes theorem is not complete yet and we need to normalize.","metadata":{}},{"cell_type":"code","source":"# prior*likelihood divided by normalizer = posterior, the normalizer can also be thought of p(seeing the data)\n#without this step our weights wouldn't be probabilities since they would not add to 1\nnormalizer = resp[['A', 'B', 'C']].sum(axis=1).to_numpy()","metadata":{"execution":{"iopub.status.busy":"2022-07-17T12:41:23.713692Z","iopub.execute_input":"2022-07-17T12:41:23.714150Z","iopub.status.idle":"2022-07-17T12:41:23.725442Z","shell.execute_reply.started":"2022-07-17T12:41:23.714107Z","shell.execute_reply":"2022-07-17T12:41:23.724119Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"resp['A'] = resp['A'] / normalizer\nresp['B'] = resp['B'] / normalizer\nresp['C'] = resp['C'] / normalizer","metadata":{"execution":{"iopub.status.busy":"2022-07-17T12:41:23.727201Z","iopub.execute_input":"2022-07-17T12:41:23.728553Z","iopub.status.idle":"2022-07-17T12:41:23.741640Z","shell.execute_reply.started":"2022-07-17T12:41:23.728498Z","shell.execute_reply":"2022-07-17T12:41:23.740115Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"resp","metadata":{"execution":{"iopub.status.busy":"2022-07-17T12:41:23.743564Z","iopub.execute_input":"2022-07-17T12:41:23.745101Z","iopub.status.idle":"2022-07-17T12:41:23.762560Z","shell.execute_reply.started":"2022-07-17T12:41:23.745042Z","shell.execute_reply":"2022-07-17T12:41:23.761240Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"### Woot! Woot! We completed the E-Step, this is really all there is, assign probabilities to each of the datapoints of coming from each cluster.\n","metadata":{}},{"cell_type":"markdown","source":"### Now compare soft assignments with the hard assignments we made earlier. In particular it is interesting to look at data point 3.\n### Originally we assigned it cluster C, but with uncertainty and variance taken into account we can see that there is actually a 1 in 4 chance it came from cluster A.\n\nAlso note that: ","metadata":{}},{"cell_type":"code","source":"resp[\"Total Assignment\"] = resp[['A', 'B', 'C']].sum(axis=1)\nresp","metadata":{"execution":{"iopub.status.busy":"2022-07-17T12:41:23.764064Z","iopub.execute_input":"2022-07-17T12:41:23.764434Z","iopub.status.idle":"2022-07-17T12:41:23.785006Z","shell.execute_reply.started":"2022-07-17T12:41:23.764400Z","shell.execute_reply":"2022-07-17T12:41:23.783773Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"The total assignments equal 1. Which means that each of the datapoints is sure to belong to one of the clusters. We are talking probabilities after all.\n\n\nAnd:","metadata":{}},{"cell_type":"code","source":"print(f\"We can simply get the weight for each cluster: \\n \\n{resp[['A', 'B', 'C']].sum(axis=0)}\")","metadata":{"execution":{"iopub.status.busy":"2022-07-17T12:41:23.786660Z","iopub.execute_input":"2022-07-17T12:41:23.786996Z","iopub.status.idle":"2022-07-17T12:41:23.795837Z","shell.execute_reply.started":"2022-07-17T12:41:23.786965Z","shell.execute_reply":"2022-07-17T12:41:23.794907Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"by summing over all the responsibilities in each cluster. \nNow given the data we have, what would be an educated guess if we'd have a new datapoint J of which we do not know its features yet?\n","metadata":{}},{"cell_type":"code","source":"total_observations = resp[['A', 'B', 'C']].sum(axis=0) #often denoted as pi for prior\nprint(f\"if we don't know better estimate with what we've seen, p(J = Cluster_k) = \\pi_k = (sum of observations in cluster K) / (sum of all observations) = \\n{total_observations / data.shape[0]}\\n\")\nprint(f\"also note that {np.sum(total_observations)} == n == num observations\")\ncluster_weights = total_observations / data.shape[0]","metadata":{"execution":{"iopub.status.busy":"2022-07-17T12:41:23.797207Z","iopub.execute_input":"2022-07-17T12:41:23.797835Z","iopub.status.idle":"2022-07-17T12:41:23.812946Z","shell.execute_reply.started":"2022-07-17T12:41:23.797801Z","shell.execute_reply":"2022-07-17T12:41:23.811712Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"### Triple Woot! these are our new priors!\nThis is updating in action, see originally we thought they all had equal chance of belonging to any cluster, but now it seems that Cluster B has more weight and in general we're less likely to see samples from cluster C.\n\n# Ok, now lets get our new parameter estimations for clusters A, B and C\nThe MLE estimator for \\mu of a univariate Gaussian distribution is simply the average of the samples. Our samples are fractions, but that won't stop us.\n\nNote: For Multivariate Normals \\mu is a vector of means.","metadata":{}},{"cell_type":"code","source":"eps = 0.0001\ndef get_mus():\n    mus = {}\n    for cluster in clusters:\n        #now we get the weighted average of all points in the cluster = 1/n*(resp*datapoint)\n        cluster_mean = np.sum([data.x * resp[cluster], data.y * resp[cluster]], axis=1)\n        if total_observations[cluster] < eps:\n            cluster_mean /= eps\n        else:\n            cluster_mean /= total_observations[cluster]\n        mus[cluster] = cluster_mean\n    return mus","metadata":{"execution":{"iopub.status.busy":"2022-07-17T12:41:23.814590Z","iopub.execute_input":"2022-07-17T12:41:23.814934Z","iopub.status.idle":"2022-07-17T12:41:23.828554Z","shell.execute_reply.started":"2022-07-17T12:41:23.814903Z","shell.execute_reply":"2022-07-17T12:41:23.826834Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"mus = get_mus()","metadata":{"execution":{"iopub.status.busy":"2022-07-17T12:41:23.830230Z","iopub.execute_input":"2022-07-17T12:41:23.830748Z","iopub.status.idle":"2022-07-17T12:41:23.843181Z","shell.execute_reply.started":"2022-07-17T12:41:23.830709Z","shell.execute_reply":"2022-07-17T12:41:23.841835Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"### Means are only half the story, next the variances","metadata":{}},{"cell_type":"code","source":"datapoints = data[['x', 'y']].to_numpy()","metadata":{"execution":{"iopub.status.busy":"2022-07-17T12:41:23.845038Z","iopub.execute_input":"2022-07-17T12:41:23.845472Z","iopub.status.idle":"2022-07-17T12:41:23.856759Z","shell.execute_reply.started":"2022-07-17T12:41:23.845436Z","shell.execute_reply":"2022-07-17T12:41:23.855535Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"def get_vars():\n    variances = {}\n    #we use the biased estimator 1/n((x_i - mu_k)*(x_i - mu_k)^T)\n    for cluster in clusters:\n        cov_matrix = np.zeros_like(cov_init)\n        #somehow the @ operator doesn't insert the missing column, I could use expand_dims but I find that less readable than np.outer(a, b) = outer product of a and b\n        for datapoint, weight in zip(datapoints, resp[cluster]):\n            #distance from mean\n            dm = (datapoint - mus[cluster])\n            cov_matrix += np.outer(dm, dm)*weight\n            if total_observations[cluster] < eps:\n                cov_matrix /= eps\n            else:\n                cov_matrix /= total_observations[cluster]\n        variances[cluster] = cov_matrix\n    return variances","metadata":{"execution":{"iopub.status.busy":"2022-07-17T12:41:23.858500Z","iopub.execute_input":"2022-07-17T12:41:23.858865Z","iopub.status.idle":"2022-07-17T12:41:23.869692Z","shell.execute_reply.started":"2022-07-17T12:41:23.858831Z","shell.execute_reply":"2022-07-17T12:41:23.868552Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"variances = get_vars()","metadata":{"execution":{"iopub.status.busy":"2022-07-17T12:41:23.871653Z","iopub.execute_input":"2022-07-17T12:41:23.872486Z","iopub.status.idle":"2022-07-17T12:41:23.882807Z","shell.execute_reply.started":"2022-07-17T12:41:23.872439Z","shell.execute_reply":"2022-07-17T12:41:23.881627Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"mus","metadata":{"execution":{"iopub.status.busy":"2022-07-17T12:41:23.884980Z","iopub.execute_input":"2022-07-17T12:41:23.885863Z","iopub.status.idle":"2022-07-17T12:41:23.896934Z","shell.execute_reply.started":"2022-07-17T12:41:23.885816Z","shell.execute_reply":"2022-07-17T12:41:23.895678Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"variances","metadata":{"execution":{"iopub.status.busy":"2022-07-17T12:41:23.899016Z","iopub.execute_input":"2022-07-17T12:41:23.899819Z","iopub.status.idle":"2022-07-17T12:41:23.912812Z","shell.execute_reply.started":"2022-07-17T12:41:23.899768Z","shell.execute_reply":"2022-07-17T12:41:23.911535Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"#these are the normal distributions we initialize with mean defined as their center and covariance matrix above\ncluster_a = multivariate_normal(mean = centroids.iloc[0], cov = cov_init)\ncluster_b = multivariate_normal(mean = centroids.iloc[1], cov = cov_init)\ncluster_c = multivariate_normal(mean = centroids.iloc[2], cov = cov_init)\nsamples = cluster_a.rvs(size = 100000)\nsamples_b =cluster_b.rvs(size = 100000)\nsamples_c =cluster_c.rvs(size = 100000)\nplt.scatter(samples[:, 0], samples[:, 1], alpha = 0.2)\nplt.scatter(samples_b[:, 0], samples_b[:, 1], alpha = 0.2)\nplt.scatter(samples_c[:, 0], samples_c[:, 1], alpha = 0.2)\nfor datapoint in datapoints:\n    plt.scatter(datapoint[0], datapoint[1], color = 'red')\nsns.set(rc={'figure.figsize':(10,10)})\nplt.show()","metadata":{"execution":{"iopub.status.busy":"2022-07-17T12:41:23.915247Z","iopub.execute_input":"2022-07-17T12:41:23.916405Z","iopub.status.idle":"2022-07-17T12:41:24.917170Z","shell.execute_reply.started":"2022-07-17T12:41:23.916331Z","shell.execute_reply":"2022-07-17T12:41:24.915578Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"#these are the normal distributions we initialize with mean defined as their center and covariance matrix above\ncluster_a = multivariate_normal(mean = mus['A'], cov =  variances['A'], allow_singular = True)\ncluster_b = multivariate_normal(mean = mus['B'], cov =  variances['B'], allow_singular = True)\ncluster_c = multivariate_normal(mean = mus['C'], cov =  variances['C'], allow_singular = True)\nsamples = cluster_a.rvs(size = 100000)\nsamples_b =cluster_b.rvs(size = 100000)\nsamples_c =cluster_c.rvs(size = 100000)\nplt.scatter(samples[:, 0], samples[:, 1], alpha = 0.2)\nplt.scatter(samples_b[:, 0], samples_b[:, 1], alpha = 0.2)\nplt.scatter(samples_c[:, 0], samples_c[:, 1], alpha = 0.2)\nfor datapoint in datapoints:\n    plt.scatter(datapoint[0], datapoint[1], color = 'red')\nsns.set(rc={'figure.figsize':(10,10)})\nplt.show()","metadata":{"execution":{"iopub.status.busy":"2022-07-17T12:41:24.918963Z","iopub.execute_input":"2022-07-17T12:41:24.919395Z","iopub.status.idle":"2022-07-17T12:41:25.858764Z","shell.execute_reply.started":"2022-07-17T12:41:24.919334Z","shell.execute_reply":"2022-07-17T12:41:25.857169Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"for i in range(1):\n    #E-Step assign datapoints\n    #likelihood * prior \n    prior = cluster_weights\n    resp['A'] = cluster_a.pdf(data[['x','y']])*cluster_weights['A']\n    resp['B'] = cluster_b.pdf(data[['x','y']])*cluster_weights['B']\n    resp['C'] = cluster_c.pdf(data[['x','y']])*cluster_weights['C']\n    total_observations = resp[['A', 'B', 'C']].sum(axis=0) \n    cluster_weights = total_observations / data.shape[0]\n    \n    #M-Step optimize the parameters\n    variances = get_vars()\n    mus = get_mus()\n    #these are the normal distributions we initialize with mean defined as their center and covariance matrix above\n    cluster_a = multivariate_normal(mean = mus['A'], cov =  variances['A'], allow_singular = True)\n    cluster_b = multivariate_normal(mean = mus['B'], cov =  variances['B'], allow_singular = True)\n    cluster_c = multivariate_normal(mean = mus['C'], cov =  variances['C'], allow_singular = True)","metadata":{"execution":{"iopub.status.busy":"2022-07-17T12:41:25.860477Z","iopub.execute_input":"2022-07-17T12:41:25.861188Z","iopub.status.idle":"2022-07-17T12:41:25.881957Z","shell.execute_reply.started":"2022-07-17T12:41:25.861154Z","shell.execute_reply":"2022-07-17T12:41:25.880713Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"samples = cluster_a.rvs(size = 100000)\nsamples_b =cluster_b.rvs(size = 100000)\nsamples_c =cluster_c.rvs(size = 100000)\nplt.scatter(samples[:, 0], samples[:, 1], alpha = 0.2)\nplt.scatter(samples_b[:, 0], samples_b[:, 1], alpha = 0.2)\nplt.scatter(samples_c[:, 0], samples_c[:, 1], alpha = 0.2)\nfor datapoint in datapoints:\n    plt.scatter(datapoint[0], datapoint[1], color = 'red')\nsns.set(rc={'figure.figsize':(10,10)})\nplt.show()","metadata":{"execution":{"iopub.status.busy":"2022-07-17T12:41:25.883495Z","iopub.execute_input":"2022-07-17T12:41:25.884036Z","iopub.status.idle":"2022-07-17T12:41:26.859339Z","shell.execute_reply.started":"2022-07-17T12:41:25.884003Z","shell.execute_reply":"2022-07-17T12:41:26.858544Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"### Possibly a fun exercise, with lots of data this will likely not work very well because everything is exponential. \nInstead maybe try the log likelihood and play with below example\nsklearns implementation is pretty straight forward: https://github.com/scikit-learn/scikit-learn/blob/baf0ea25d/sklearn/mixture/_gaussian_mixture.py#L456\nor you know, instead of reinventing the wheel just use it :-) knowing now with confidence that you can compute it by hand if needed","metadata":{}},{"cell_type":"code","source":"cov_a = np.array([[0.5, -1.5], \n                  [-1.5, 5.5]]) #long\nmean_a = np.array([8.5, 5.0])\ncov_b = np.array([[1, 0.0], \n                  [0.0, 1.0]]) #circle\nmean_b = np.array([2, 2])\ncov_c = np.array([[5.5, 0.8], \n                  [0.8, 0.8]]) #wide\nmean_c = np.array([1.5, 5.0])\n\n\na = multivariate_normal(mean_a, cov_a)\nb = multivariate_normal(mean_b, cov_b)\nc = multivariate_normal(mean_c, cov_c)\nsize_a, size_b, size_c = 700, 100, 200\nsamples_a = a.rvs(size = size_a) #a should get a much higher weight after some time\nsamples_b =b.rvs(size = size_b)\nsamples_c =c.rvs(size = size_c)\ndata = np.concatenate([samples_a, samples_b, samples_c])\n# data = np.array([[10.0, 5.0], [2.0, 1], [3, 7], [5, 6], [9, 4], [3, 1], [2, 3], [3, 5], [5, 5]])\n# data = np.array([[10.0, 5.0], [2.0, 1], [3, 7]])\ndata = pd.DataFrame(data, columns = ['x', 'y'])\ndata['colors'] = ['red' for i in range(size_a)] + ['blue' for i in range(size_b)] + ['yellow' for i in range(size_c)]\nsns.set(rc={'figure.figsize':(10,10)})\nplt.scatter(data.x, data.y, s = 150, c = data['colors'])\n# plt.scatter(data.x, data.y, s = 150)\nplt.show()","metadata":{"execution":{"iopub.status.busy":"2022-07-17T12:41:26.864302Z","iopub.execute_input":"2022-07-17T12:41:26.864943Z","iopub.status.idle":"2022-07-17T12:41:27.174593Z","shell.execute_reply.started":"2022-07-17T12:41:26.864907Z","shell.execute_reply":"2022-07-17T12:41:27.173519Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"from sklearn.mixture import GaussianMixture\ngmm = GaussianMixture(n_components = 3, verbose = 1, max_iter = 10) #set max_iter higher or use basian gaussian mixture with unknown number of components\ngmm.fit(data[['x', 'y']])","metadata":{"execution":{"iopub.status.busy":"2022-07-17T12:41:27.176340Z","iopub.execute_input":"2022-07-17T12:41:27.177052Z","iopub.status.idle":"2022-07-17T12:41:27.862085Z","shell.execute_reply.started":"2022-07-17T12:41:27.177008Z","shell.execute_reply":"2022-07-17T12:41:27.860569Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"data[['A', 'B', 'C']] = np.around(gmm.predict_proba(data[['x', 'y']]), 2)\ndata","metadata":{"execution":{"iopub.status.busy":"2022-07-17T12:41:27.864441Z","iopub.execute_input":"2022-07-17T12:41:27.865367Z","iopub.status.idle":"2022-07-17T12:41:27.904449Z","shell.execute_reply.started":"2022-07-17T12:41:27.865305Z","shell.execute_reply":"2022-07-17T12:41:27.903142Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"gmm.means_","metadata":{"execution":{"iopub.status.busy":"2022-07-17T12:41:27.906696Z","iopub.execute_input":"2022-07-17T12:41:27.907643Z","iopub.status.idle":"2022-07-17T12:41:27.917477Z","shell.execute_reply.started":"2022-07-17T12:41:27.907591Z","shell.execute_reply":"2022-07-17T12:41:27.916173Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"gmm.covariances_","metadata":{"execution":{"iopub.status.busy":"2022-07-17T12:41:27.920066Z","iopub.execute_input":"2022-07-17T12:41:27.921173Z","iopub.status.idle":"2022-07-17T12:41:27.937190Z","shell.execute_reply.started":"2022-07-17T12:41:27.921121Z","shell.execute_reply":"2022-07-17T12:41:27.935730Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"#priors our cluster weights from before\ngmm.weights_","metadata":{"execution":{"iopub.status.busy":"2022-07-17T12:41:27.939960Z","iopub.execute_input":"2022-07-17T12:41:27.941035Z","iopub.status.idle":"2022-07-17T12:41:27.955408Z","shell.execute_reply.started":"2022-07-17T12:41:27.940979Z","shell.execute_reply":"2022-07-17T12:41:27.953836Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Actually sklearn didn't do so well 🤣\nGood reminder that this happens from time to time, it doesn't have to find the global maximum but can converge to a local maximum, EM can easily overfit, to see this look at the normal distributions formula and think about what happens to likelihood N(x | mu, sigma) as sigma -> 0\n\nMost of the times it does quite well though.\n","metadata":{}},{"cell_type":"code","source":"plt.scatter(data.x, data.y, s = 150, c = data['colors'])\nplt.scatter(gmm.means_[:, 0], gmm.means_[:, 1], s = 150, color = 'black')\nplt.show()","metadata":{"execution":{"iopub.status.busy":"2022-07-17T12:41:27.958114Z","iopub.execute_input":"2022-07-17T12:41:27.959181Z","iopub.status.idle":"2022-07-17T12:41:28.272658Z","shell.execute_reply.started":"2022-07-17T12:41:27.959126Z","shell.execute_reply":"2022-07-17T12:41:28.271520Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"#I now also added kaggle's july tabular playground \n#where you can actually try and practice this with a real competition dataset\n#have fun!","metadata":{},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"dataset = pd.read_csv('../input/tabular-playground-series-jul-2022/data.csv')","metadata":{},"execution_count":null,"outputs":[]}]}