{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Model selection using GpABC\n",
    "\n",
    "## Problem formulation\n",
    "\n",
    "This notebook illustrates how to perform [model selection](https://tanhevg.github.io/GpABC.jl/latest/overview-ms/) using the algorithm described in (Toni et al, 2009). The example here is adapted from the SIR (susceptible-infected-recovered) model in Section 3.3 of the [paper](https://royalsocietypublishing.org/doi/10.1098/rsif.2008.0172#d3e1991).\n",
    "\n",
    "The example gives two models that describe the evolution of the number of of susceptible ($S$), infected ($I$) and recovered ($R$) individuals in a population (the number of dead individuals is not included in the model).\n",
    "\n",
    "There are two possible models, both of which are described below. Given some observed data on the number of individuals in each disease state over time we wish to determine which of the two models is better supported by these data. The data was generated using the first model so the algorithm should reflect this.\n",
    "\n",
    "### Model 1\n",
    "\n",
    "A basic SIR model assumes that each individual can be infected once and is then able to infect other individuals immediately. \n",
    "\n",
    "The model is given by\n",
    "\n",
    "\\begin{align}\n",
    "    \\dot{S} &= \\alpha - \\gamma S I - d S, \\\\\n",
    "    \\dot{I} &= \\gamma S I - v I - d I, \\\\\n",
    "    \\dot{R} &= v I - d R \\, ,\n",
    "\\end{align}\n",
    "\n",
    "where the notation $\\dot{x}$ denotes $\\text{d}x/\\text{d}t$.\n",
    "\n",
    "The paramters of this model are the birth rate $\\alpha$, the infection rate $\\gamma$, the recovery rate $v$ and the death rate $d$. Note that the death rate is constant for all three disease states.\n",
    "\n",
    "### Model 2\n",
    "\n",
    "A slightly more sophisticated model includes a time delay between an individual becoming infected and being able to infect others by introducing an additional latent state ($L$). This model is given by\n",
    "\n",
    "\\begin{align}\n",
    "    \\dot{S} &= \\alpha - \\gamma S I - d S , \\\\\n",
    "    \\dot{L} &= \\gamma S I - \\delta L - d L , \\\\\n",
    "    \\dot{I} &= \\delta L - v I - d I , \\\\\n",
    "    \\dot{R} &= v I - d R \\, ,\n",
    "\\end{align}\n",
    "\n",
    "where $\\delta$ is the transition rate from latent to infected.\n",
    "\n",
    "### References\n",
    "- Toni, T., Welch, D., Strelkowa, N., Ipsen, A. and Stumpf, M.P., 2009. Approximate Bayesian computation scheme for parameter inference and model selection in dynamical systems. Journal of the Royal Society Interface, 6(31), pp.187-202. https://doi.org/10.1098/rsif.2008.0172\n",
    "\n",
    "# To run this notebook locally:\n",
    "* Make sure that [Jupyter](https://jupyter.org/) is installed, along with [Julia](https://www.julialang.org/) and its [Jupyter Kernel](https://github.com/JuliaLang/IJulia.jl).\n",
    "* Clone or download [GpABC.jl](https://github.com/tanhevg/GpABC.jl) to your machine.\n",
    "* Run `jupyter` from the `examples` directory of the local copy of `GpABC.jl`:\n",
    "```bash\n",
    "$ cd GpABC.jl/examples\n",
    "$ jupyter notebook model-selection-example.ipynb\n",
    "```\n",
    "* The first line of the cell below activates the local project environment and downloads all the dependencies.\n",
    "\n",
    "This notebook was tested under Julia 1.7.2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[32m\u001b[1m  Activating\u001b[22m\u001b[39m project at `~/gaussian_processes/GpABC.jl/examples`\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "Plots.PyPlotBackend()"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ENV[\"PYTHON\"]=\"\"; import Pkg; Pkg.activate(@__DIR__); Pkg.instantiate()\n",
    "using GpABC, OrdinaryDiffEq, Distributions, Distances, Plots\n",
    "pyplot()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Prepare the observed data into the correct format: a 2D array with size (number of trajectories, number of time points).\n",
    "\n",
    "The observed data contains information the number of susceptible, infected and recovered individuals at 13 time points."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "times = [0.0, 0.6, 1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 7.0, 8.0, 9.0, 10.0, 11.0]\n",
    "data = [[ 20.     ,  10.     ,   0.     ],\n",
    "        [  0.12313,  13.16813,   9.42344],\n",
    "        [  0.12102,   7.17251,  11.18957],\n",
    "        [  0.09898,   2.36466,  10.0365 ],\n",
    "        [  0.37887,   0.92019,   6.87117],\n",
    "        [  1.00661,   0.61958,   4.44955],\n",
    "        [  1.20135,   0.17449,   3.01271],\n",
    "        [  1.46433,   0.28039,   1.76431],\n",
    "        [  1.37789,   0.0985 ,   1.28868],\n",
    "        [  1.57073,   0.03343,   0.81813],\n",
    "        [  1.4647 ,   0.28544,   0.52111],\n",
    "        [  1.24719,   0.10138,   0.22746],\n",
    "        [  1.56065,   0.21671,   0.19627]]\n",
    "\n",
    "data = hcat(data...); "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Next, we define the models. As in other `GpABC` functions, the model is represented using a function that takes parameters as an argument and returns trajectories that are the solution to the differential equation system. This is called the `simulator_function`. The solution should have the same shape as the reference data, which is (number of trajectories, number of time points).\n",
    "\n",
    "This means that for the second model, which has 4 states, we need to modify the `simulator_function` in order to ignore the solution for the latent state $L$ as we have no reference data for that trajectory. \n",
    "\n",
    "We also need to define different initial conditions for the two models as the first model has 3 states and the second has 4."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Need initial conditions for each model since they have different numbers of states\n",
    "ic1 = [20.0, 10.0, 0.0]\n",
    "ic2 = [20.0, 0.0, 10.0, 0.0]\n",
    "ics = [ic1, ic2]\n",
    "\n",
    "# Define a simulator function for each model. Each function returns the solution of the model\n",
    "# as a 2D array for some parameter values. The appropriate ODE is solved at the same time \n",
    "# points as the observed data\n",
    "function simulator1(params::Array{Float64,1})\n",
    "    \n",
    "    # p = (alpha, gamma, d, v)\n",
    "    # x = (S, I, R)\n",
    "    function model1(dx, x, p, t)\n",
    "        dx[1] = p[1] - p[2]*x[1]*x[2] - p[3]*x[1] # dS/dt = alpha - gamma*S*I - d*S\n",
    "        dx[2] = p[3]*x[1]*x[2] - p[4]*x[2] - p[3]*x[2] # dI/dt = gamma*S*I - v*I - d*I\n",
    "        dx[3] = p[4]*x[2] - p[3]*x[3] # dR/dt = v*I - d*R\n",
    "    end\n",
    "    \n",
    "    sol = solve(ODEProblem(model1, ics[1], (times[1], times[end]), params),\n",
    "                RK4(), saveat=times, force_dtmin=true)\n",
    "    hcat(sol.u...)\n",
    "end\n",
    "\n",
    "function simulator2(params::Array{Float64,1})\n",
    "\n",
    "    # p = (alpha, gamma, d, v, delta)\n",
    "    # x = (S, L, I, R)\n",
    "    function model2(dx, x, p, t)\n",
    "        dx[1] = p[1] - p[2]*x[1]*x[3] - p[3]*x[1] # dS/dt = alpha - gamma*S*I - d*S\n",
    "        dx[2] = p[2]*x[1]*x[3] - p[5]*x[2] - p[3]*x[2] # dL/dt = gamma*S*I - delta*L - d*L\n",
    "        dx[3] = p[5]*x[2] - p[4]*x[3] - p[3]*x[3] # dI/dt = delta*L - v*I - d*I\n",
    "        dx[4] = p[4]*x[3] - p[3]*x[4] # dR/dt = v*I - d*R\n",
    "    end\n",
    "    \n",
    "    # Model2 contains the species L, which is not measured - we remove it from the returned ODE solution\n",
    "    # so that it can be compared to the reference data \"data\", which only contains S, I and R\n",
    "    sol = solve(ODEProblem(model2, ics[2], (times[1], times[end]), params),\n",
    "                RK4(), saveat=times, force_dtmin=true)\n",
    "    hcat(sol.u...)[[1,3,4],:]\n",
    "end;"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We also need a set of priors for each model, as well as a summary statistic and threshold schedule. We will compare the full model solution with the data, which is the \"`keep_all`\" summary statistic. The model prior will be the default discrete uniform prior with span 2 (as there are two models)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "#\n",
    "# Priors and initial conditions - these are model-specfic as each model can \n",
    "# have different numbers of parameters/species\n",
    "#\n",
    "priors1 = [Uniform(0.0, 5.0) for i in 1:4]\n",
    "priors2 = vcat([Uniform(0.0, 5.0) for i in 1:4], Uniform(0.0, 10.0))\n",
    "priors3 = vcat([Uniform(0.0, 5.0) for i in 1:4], Uniform(0.0, 10.0))\n",
    "\n",
    "threshold_schedule = [20, 15, 10, 5, 3, 2.5, 2, 1.7, 1.5]\n",
    "summary_statistic = \"keep_all\";"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We are now ready to perform the model selection computation by calling [`SimulatedModelSelection`](https://tanhevg.github.io/GpABC.jl/latest/ref-ms/#GpABC.SimulatedModelSelection) and accepting 200 particles at each population (`n_particles = 200`).\n",
    "\n",
    "First, we will do this using full model simulations:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "┌ Info: GpABC model selection simulation  Population 1 - ABC Rejection ϵ = 20.0\n",
      "└ @ GpABC /data/.julia/packages/GpABC/o0EN1/src/abc/model_selection.jl:9\n",
      "┌ Info: GpABC model selection simulation\n",
      "│ Completed 712 iterations, accepting 200 particles in total.\n",
      "│ Number of accepted parameters by model:\n",
      "│ Model 1: 113\tModel 2: 87\n",
      "└ @ GpABC /data/.julia/packages/GpABC/o0EN1/src/abc/model_selection.jl:113\n",
      "┌ Info: GpABC model selection simulation  Population 2 - ABC SMC ϵ = 15.0\n",
      "└ @ GpABC /data/.julia/packages/GpABC/o0EN1/src/abc/model_selection.jl:24\n",
      "┌ Info: GpABC model selection simulation\n",
      "│ Completed 1406 iterations, accepting 200 particles in total.\n",
      "│ Number of accepted parameters by model: \n",
      "│ Model 1: 42\tModel 2: 158\n",
      "└ @ GpABC /data/.julia/packages/GpABC/o0EN1/src/abc/model_selection.jl:307\n",
      "┌ Info: GpABC model selection simulation  Population 3 - ABC SMC ϵ = 10.0\n",
      "└ @ GpABC /data/.julia/packages/GpABC/o0EN1/src/abc/model_selection.jl:24\n",
      "┌ Info: GpABC model selection simulation\n",
      "│ Completed 2702 iterations, accepting 200 particles in total.\n",
      "│ Number of accepted parameters by model: \n",
      "│ Model 1: 19\tModel 2: 181\n",
      "└ @ GpABC /data/.julia/packages/GpABC/o0EN1/src/abc/model_selection.jl:307\n",
      "┌ Info: GpABC model selection simulation  Population 4 - ABC SMC ϵ = 5.0\n",
      "└ @ GpABC /data/.julia/packages/GpABC/o0EN1/src/abc/model_selection.jl:24\n",
      "┌ Info: GpABC model selection simulation\n",
      "│ Completed 6656 iterations, accepting 200 particles in total.\n",
      "│ Number of accepted parameters by model: \n",
      "│ Model 1: 21\tModel 2: 179\n",
      "└ @ GpABC /data/.julia/packages/GpABC/o0EN1/src/abc/model_selection.jl:307\n",
      "┌ Info: GpABC model selection simulation  Population 5 - ABC SMC ϵ = 3.0\n",
      "└ @ GpABC /data/.julia/packages/GpABC/o0EN1/src/abc/model_selection.jl:24\n",
      "┌ Info: GpABC model selection simulation\n",
      "│ Completed 5800 iterations, accepting 200 particles in total.\n",
      "│ Number of accepted parameters by model: \n",
      "│ Model 1: 113\tModel 2: 87\n",
      "└ @ GpABC /data/.julia/packages/GpABC/o0EN1/src/abc/model_selection.jl:307\n",
      "┌ Info: GpABC model selection simulation  Population 6 - ABC SMC ϵ = 2.5\n",
      "└ @ GpABC /data/.julia/packages/GpABC/o0EN1/src/abc/model_selection.jl:24\n",
      "┌ Info: GpABC model selection simulation\n",
      "│ Completed 1544 iterations, accepting 200 particles in total.\n",
      "│ Number of accepted parameters by model: \n",
      "│ Model 1: 165\tModel 2: 35\n",
      "└ @ GpABC /data/.julia/packages/GpABC/o0EN1/src/abc/model_selection.jl:307\n",
      "┌ Info: GpABC model selection simulation  Population 7 - ABC SMC ϵ = 2.0\n",
      "└ @ GpABC /data/.julia/packages/GpABC/o0EN1/src/abc/model_selection.jl:24\n",
      "┌ Info: GpABC model selection simulation\n",
      "│ Completed 1986 iterations, accepting 200 particles in total.\n",
      "│ Number of accepted parameters by model: \n",
      "│ Model 1: 127\tModel 2: 73\n",
      "└ @ GpABC /data/.julia/packages/GpABC/o0EN1/src/abc/model_selection.jl:307\n",
      "┌ Info: GpABC model selection simulation  Population 8 - ABC SMC ϵ = 1.7\n",
      "└ @ GpABC /data/.julia/packages/GpABC/o0EN1/src/abc/model_selection.jl:24\n",
      "┌ Info: GpABC model selection simulation\n",
      "│ Completed 2340 iterations, accepting 200 particles in total.\n",
      "│ Number of accepted parameters by model: \n",
      "│ Model 1: 103\tModel 2: 97\n",
      "└ @ GpABC /data/.julia/packages/GpABC/o0EN1/src/abc/model_selection.jl:307\n",
      "┌ Info: GpABC model selection simulation  Population 9 - ABC SMC ϵ = 1.5\n",
      "└ @ GpABC /data/.julia/packages/GpABC/o0EN1/src/abc/model_selection.jl:24\n",
      "┌ Info: GpABC model selection simulation\n",
      "│ Completed 1912 iterations, accepting 200 particles in total.\n",
      "│ Number of accepted parameters by model: \n",
      "│ Model 1: 119\tModel 2: 81\n",
      "└ @ GpABC /data/.julia/packages/GpABC/o0EN1/src/abc/model_selection.jl:307\n"
     ]
    }
   ],
   "source": [
    "n_particles = 200\n",
    "\n",
    "ms_sim_result = SimulatedModelSelection(data,\n",
    "    [simulator1, simulator2],\n",
    "    [priors1, priors2],\n",
    "    threshold_schedule,\n",
    "    n_particles);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "A plot recipe allows this result to be plotted easily:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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"
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "plot(ms_sim_result, title=\"Model selection result - simulation\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The number of accepted particles per population can be seen in the plot. Here, the \"population\" is the algorithmic population approach that is also taken during ABC-SMC. The number of particles accepted may be taken as a proxy for the posterior probability. Hence, the first model is correctly identified as the correct one.\n",
    "\n",
    "Now, do the same computation using emulation by calling [`EmulatedModelSelection`](https://tanhevg.github.io/GpABC.jl/latest/ref-ms/#GpABC.EmulatedModelSelection). The only difference between the simulation and emulation function call is that we specify the number of design points as the first argument when using emulation.\n",
    "\n",
    "We will use 200 design points."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "┌ Info: GpABC model selection emulation  Population 1 - ABC Rejection ϵ = 20.0\n",
      "└ @ GpABC /data/.julia/packages/GpABC/o0EN1/src/abc/model_selection.jl:9\n",
      "┌ Info: GpABC model selection emulation\n",
      "│ Completed 15 iterations, accepting 200 particles in total.\n",
      "│ Number of accepted parameters by model: \n",
      "│ Model 1: 109\tModel 2: 91\n",
      "└ @ GpABC /data/.julia/packages/GpABC/o0EN1/src/abc/model_selection.jl:212\n",
      "┌ Info: GpABC model selection emulation  Population 2 - ABC SMC ϵ = 15.0\n",
      "└ @ GpABC /data/.julia/packages/GpABC/o0EN1/src/abc/model_selection.jl:24\n",
      "┌ Info: GpABC model selection emulation\n",
      "│ Completed 55 iterations, accepting 200 particles in total.\n",
      "│ Number of accepted parameters by model: \n",
      "│ Model 1: 21\tModel 2: 179\n",
      "└ @ GpABC /data/.julia/packages/GpABC/o0EN1/src/abc/model_selection.jl:398\n",
      "┌ Info: GpABC model selection emulation  Population 3 - ABC SMC ϵ = 10.0\n",
      "└ @ GpABC /data/.julia/packages/GpABC/o0EN1/src/abc/model_selection.jl:24\n",
      "┌ Info: GpABC model selection emulation\n",
      "│ Completed 80 iterations, accepting 200 particles in total.\n",
      "│ Number of accepted parameters by model: \n",
      "│ Model 1: 9\tModel 2: 191\n",
      "└ @ GpABC /data/.julia/packages/GpABC/o0EN1/src/abc/model_selection.jl:398\n",
      "┌ Info: GpABC model selection emulation  Population 4 - ABC SMC ϵ = 5.0\n",
      "└ @ GpABC /data/.julia/packages/GpABC/o0EN1/src/abc/model_selection.jl:24\n",
      "┌ Info: GpABC model selection emulation\n",
      "│ Completed 62 iterations, accepting 200 particles in total.\n",
      "│ Number of accepted parameters by model: \n",
      "│ Model 1: 200\tModel 2: 0\n",
      "└ @ GpABC /data/.julia/packages/GpABC/o0EN1/src/abc/model_selection.jl:398\n",
      "┌ Warning: All but one model is dead after population 4 - terminating model selection algorithm\n",
      "└ @ GpABC /data/.julia/packages/GpABC/o0EN1/src/abc/model_selection.jl:35\n"
     ]
    }
   ],
   "source": [
    "ms_emu_result = EmulatedModelSelection(data,\n",
    "    [simulator1, simulator2],\n",
    "    [priors1, priors2],\n",
    "    threshold_schedule,\n",
    "    n_particles,\n",
    "    200;\n",
    "    summary_statistic = \"keep_all\",\n",
    "    distance_function=Distances.euclidean);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAlgAAAGQCAYAAAByNR6YAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8qNh9FAAAACXBIWXMAAA9hAAAPYQGoP6dpAACGDklEQVR4nO3dd3gUZdfH8e/sbnrZdHoIvUoJEHrvvSqKIE2aAlIVFSmi2EAB0RcsiAjyiPQivSsQunSkSuikhxSS3Z33D57kIZANCSQ7KedzXVyQmdnd3+4Om7P33HNGUVVVRQghhBBCZBmd1gGEEEIIIfIaKbCEEEIIIbKYFFhCCCGEEFlMCiwh/mvhwoUoisLChQuf634URaFJkyZZkulZ7dq1C0VRmDJliqY5kmXVa5vfyOuWOU2aNEFRlFz/GCJvkAJLaObq1asoioKiKBQpUgSz2ZzmdidPnkzZrnz58jZOKTIi+b3s16+f1lHyhZxQxOdVU6ZMQVEUdu3apXUUkctJgSU0ZzAYuHnzJps3b05z/Y8//ojBYLBxKpGVunbtytmzZ+natavWUYR4LosWLeLs2bNaxxC5gBRYQnP16tXDaDSyYMGCJ9YlJiayZMkS2rVrp0EykVWMRiPly5fHaDRqHUWI5+Lv7y8j6SJDpMASmnNycqJnz56sW7eO0NDQVOvWrl1LaGgo/fv3t3r7uLg4pkyZQvny5XF0dMTLy4v27duzb9++NLcPDw9n6NChFChQAGdnZ2rVqsWqVavSzXjixAlefvllChUqhL29PcWLF2fEiBGEhYVl/gk/IiEhgZkzZ1K1alWMRiOurq6UKlWKV155hZMnTz6x/Zo1a2jevDmenp44OjpSuXJlZsyYYfXwalru3r3L6NGjKV26NA4ODvj4+NC9e3dOnTpldftx48ZRrly5lNe3Tp06zJw5E3g4T6hEiRIA/PzzzymHcx89zJLeXKJ9+/bRvn17vLy8cHR0pHz58kyZMoW4uLgntk0+NHbv3j0GDBiAn58fTk5O1KlTJ1OHdPr164eiKFy+fJmvvvqKSpUq4eDgkOoQZ2ZepwsXLtC/f39KlCiBo6MjPj4+BAYGMnbs2FTbBQQEEBAQkGamjMztSZ5bB7B79+5Ur3V2ztO6cuUKr7/+Ov7+/jg4OFCoUCH69evHv//++8S2ye/RjRs36NWrFz4+Pri5udG+fXsuX74MwPnz5+natSteXl64ubnx4osvcvfu3TSfa1rzCDNzSDoqKorPPvuMxo0bU7hwYezt7SlcuDCvvfYaly5dSrVtkyZNmDp1KgBNmzZNeW0ffc+svU8mk4mvvvqKqlWr4uTkhNFopGnTpmzYsOGJbR/9/7B9+3YaNGiAi4sL3t7e9O3b97k/V0TOIMddRI4wYMAAvvvuO5YsWcJbb72VsnzBggX4+fnRoUOHNG/34MEDmjdvzoEDBwgMDGTUqFHcvXuX3377jS1btvDbb7/RrVu3lO3j4uJo0qQJJ0+epG7dujRu3JiQkBB69uxJq1at0nyMtWvX8tJLL6HX6+nUqRPFihXjzJkzzJ07l82bNxMcHIynp+czPe++ffuybNkyqlSpQv/+/XFwcODatWvs3LmT1q1b88ILL6Rs+9577/HJJ59QtGhRunfvjru7O3v27GH8+PEEBwfz+++/P/XxLl26lPLLr1WrVnTp0oW7d++yYsUKNm/ezPbt26ldu3bK9hcuXKBp06bcuHGDBg0a0KVLF2JjYzl16hQff/wxY8eOpVq1arz11lvMnj2bqlWr0qVLl5TbWysmkq1YsYKXX34Ze3t7evbsiZ+fH9u2bWPq1Kls2bKFnTt34uDgkOo2kZGR1K9fH3d3d1599dWU97t169YcOXKEypUrZ+zFB0aMGMGBAwdo3749HTp0oECBApl+nW7evElQUBCxsbG0b9+enj17cv/+fS5cuMDXX3+dUohmhYCAACZPnszUqVMpXrx4qgKjWrVqWfY4jwoODqZ169bExsbSsWNHSpcuzdWrV1myZAkbN25k//79lCxZMtVtIiIiaNCgAQULFqRv3778888/rF+/nnPnzrF27VoaNmxIYGAgAwYM4MiRIyxfvpzIyEi2bt2a5fnPnj3LpEmTaNq0KV27dsXFxYVz587x66+/smHDBo4ePUrx4sUBUl7P3bt307dv35T918PDI93HUFWVnj17snLlSsqWLcubb75JbGwsy5Yto0OHDsyePZuRI0c+cbt169axfv16OnbsyLBhw9izZw+LFi3i0qVL/Pnnn1n5MggtqEJo5MqVKyqgtm7dWlVVVa1UqZJapUqVlPXXr19X9Xq9OnbsWFVVVRVQy5Url+o+PvzwQxVQX331VdVisaQs//vvv1UHBwfV09NTjY6OTlk+efJkFVAHDRqU6n42b96sAiqg/vTTTynLQ0NDVXd3d7Vo0aLqv//+m+o2v/76qwqow4cPT7UcUBs3bvzU5x8ZGakqiqLWrFlTNZlMqdaZTCY1IiIi5ectW7aogNq2bVs1NjY2ZbnFYlGHDh2qAury5ctTlu/cuVMF1MmTJ6e633r16qkGg0HdsmVLquXnz59X3dzc1BdeeCHV8qCgIBVQv/vuuyfyh4SEpPw7+b3s27dvms/1p59+euK1jY6OVj08PFQHBwf177//TvWcevXqpQLqtGnTUt1P8nv0xhtvqGazOWX5Dz/8oALqkCFD0nz8x/Xt21cF0nxfVTVzr9OcOXNUQJ09e/YT93Pv3r1UPxcvXlwtXrx4mpkaN26sPv6RnNbrpqoZ38eeV2JiohoQEKC6ubmpx48fT7Vu7969ql6vVzt06PBENkAdPXp0quXJ+6mHh4c6a9aslOUWi0Vt166dCqhHjx5NWW5tH1ZV6/tbWq9hZGSkGhYW9sR97NixQ9XpdOrrr7+eannyZ8TOnTufuI21x1i0aFHKe/LgwYOU5SEhIaqfn59qZ2enXr58OWV58vtqMBjUP//8M2W5yWRSmzRpogLq/v3703x8kXvIIUKRY/Tv358TJ05w5MgR4OEwutlsZsCAAVZvs3DhQuzs7Pj0009TDdtXqVKFfv36ERERwZo1a1KWL1q0CHt7ez788MNU99OqVSuaN2/+xP0vWrSI6OhoPvnkE/z9/VOte+WVVwgMDOQ///nPMz1fRVFQVRUHBwf0en2qdXq9PtW35rlz5wIwf/58nJ2dU91H8nNfunRpuo937Ngx9u3bR9++fWnZsmWqdWXLlmXQoEGcPHky5RDYoUOHOHjwII0aNWLQoEFP3F/RokUz9Xwft3r1aiIjIxkwYABVqlR54jkZDIY0D3u5uLjw2WefodP97+Orb9++GAwGDh06lKkM48ePf+J9zezrlMzJyemJ+/fx8clUnpxm/fr1XL16lbfffpuqVaumWtegQQM6d+7MH3/8QXR0dKp1rq6uTJs2LdWyXr16AeDt7Z1qNEdRFF5++WUA/v777yx/DkajES8vryeWN23alEqVKrFt27bnfozk/fTzzz/H3t4+ZXnRokUZPXo0SUlJLFmy5Inb9erVi/r166f8rNfr6du3L0Cm92WR88ghQpFj9OnTh3fffZcFCxZQo0YNFi5cSO3atalYsWKa20dHR3P58mUqVKiQ5i/7Jk2aMH/+fI4fP07v3r2JiYnhypUrVKxYkYIFCz6xfcOGDdm+fXuqZQcOHEj5++LFi0/cJiEhgdDQUEJDQzP9y9Td3Z02bdqwadMmAgMD6dGjBw0bNqR27dqpPqSTH9/FxYUff/wxzftycnLi3Llz6T5e8nO5fft2mvNakm9/7tw5KleuzMGDBwGsHjp9XseOHQNIs91AsWLFKFWqFOfPnycmJgY3N7eUdWXKlMHV1TXV9gaDgQIFChAZGZmpDEFBQU8sy+zr1KFDByZMmMCbb77J1q1badOmDQ0aNKBs2bKZypLd0nouo0aNSvfwV/Jrce7cuTRvf/v2bSwWC//88w81a9ZMWV6mTBlcXFxSbVuoUCHg4Zefx+cwJa+7ceNGRp5Kpu3atYtZs2YRHBxMaGgoJpMpZd3j/9eexbFjx3Byckpzf0rev48fP/7EusDAwCeWJX+WZXZfFjmPFFgix/Dz86Ndu3YsXbqUTp06cfHiRcaNG2d1++RvzcnzZh6XXERFRUWl+tvPzy/N7dO6n/DwcAC++eabdLPHxsY+02jF8uXLmT59OkuXLuX9998HwM3NjQEDBjB9+vSU0arw8HBMJlPKBFxrGdKT/Fw2bNiQ5sTbx+8n+QO+SJEiGX4+mZGR9+/8+fNER0enKrCsnYloMBgyNdnf2mNn9nUqUaIE+/fvZ+rUqWzcuDFlLly5cuWYNm0aL774YqYyZZe09p1+/fqlW2AlvxZpjb486vF9z93d/YltklutpLcuKSkp3cd5Fr///js9e/bE1dWV1q1bExAQgLOzc8ok87Qm6mdWdHQ0xYoVS3Pd459Dj0prX05+LTK7L4ucRwoskaMMGDCANWvWMHDgQJycnHjllVesbpv8QX3nzp001ycvT94u+e/Hz1Z6fPu0HuPkyZOZmjydUS4uLnz88cd8/PHHXLlyhZ07dzJv3jxmz55NfHw88+fPT8mhKMoTZ1lmRvJz+frrrxk+fPhTt0/+xZtdowqZff+yQ1png2X2dYKHozIrVqwgKSmJI0eOsHHjRubMmUPPnj0pXLhwymEgnU5HYmJimveR1i/grKSqaqZvk/xarFu3zuqJJtkl+RDwo6NNyTLzWk2ZMgVHR0eOHDlCmTJlUq171sP7j3N3d9d0PxY5k8zBEjlKu3btKFiwIDdu3Eg5U84ad3d3SpYsycWLF9MsAnbv3g387+wqd3d3SpQowcWLF7l9+/YT2+/du/eJZclniu3fv/9Znk6mlChRggEDBrB7925cXV1Zu3ZtqhxhYWFcuHDhme8/s88l+XDHli1bnrpt8hyyzHzrrl69OkCa7RVu3LjBpUuXKFmyZKrRK1t4nvfczs6OOnXqMHXqVObMmYOqqqxfvz5lvaenJ3fv3n2iaIiNjc3Ue6vT6WwywmHL/f9xyWfmpvV/O/nwckZcunSJChUqPFFc3bx584k2DfDs+3J8fHzKYfVHPf45JPIPKbBEjmIwGFi7di2rVq3i448/fur2ffv2JSkpiXfffTfVN/RTp07x008/YTQaU7UN6NOnD4mJiUyaNCnV/WzZsuWJ+VfwcOK9m5sb77//PqdPn35ifVxcXMo8lcy6d+9emh/IERERPHjwINWk6eRJwQMGDEizR87t27ef2l06KCiI2rVrs3TpUn777bcn1lsslpRfBgC1atUiKCiIPXv28P333z+x/aO/+Dw9PVEUhevXr6eb4VGdO3fGaDTy008/pXptVVXl3XffJSkpSZNL72T2dTp06FCao6LJIxePvo81a9Z8YsJz8vN92iHeR3l5eWXqtX5WnTt3xt/fny+//JI9e/Y8sT4pKSnb2gmUK1cu5YtG8qFKePi6fvTRRxm+n+LFi3Px4sVUI0wJCQkMGzYszdGx5AnxmXl9kyemJ++3yW7cuMGXX36JwWDg1VdfzfD9ibxBDhGKHKdWrVrUqlUrQ9u+/fbbbNiwgV9++YWzZ8/SvHlz7t27x2+//UZSUhKLFi1KNQLy9ttvs3LlSr7//ntOnz5No0aNCAkJYdmyZbRv3/6JOTe+vr4sXbqUF198kapVq9KmTRvKly9PQkIC//77L7t376ZevXps2rQp08/zxo0b1K5dm0qVKhEYGEiRIkUICwtjzZo1JCUl8fbbb6ds26ZNGz744AOmTZtG6dKladOmDcWLFycsLIyLFy+yd+9ePvroIypUqJDuYy5dupSmTZvy8ssvM2vWLGrUqIGjoyPXrl1j//793Lt3j4SEhJTtFy9eTJMmTRg8eDC//PILdevWJSEhgdOnT3Ps2LGUYs/V1ZVatWqxZ88e+vfvT5kyZdDpdPTq1euJs/SSubu78/333/PKK69Qu3Ztevbsia+vL9u3b+fw4cMEBQUxfvz4TL+uWSEzr9OSJUv49ttvadKkCaVLl8bd3Z0zZ87wxx9/4OPjk+os2OHDh/PTTz/x+uuvs3XrVnx9fdm7dy+RkZFUrVo1w2fRNWvWjGXLltGjRw+qV6+OXq+nffv2qfqmZQUHBweWL19O27Ztady4Mc2bN085VH7t2jX27t2Lt7f3U0+weBb29vYMHz6cTz/9lMDAQDp37kxMTAzr1q2jcePGaY4+pWXEiBGMGDGC6tWr06NHD0wmE1u3bkVV1TRf8+QGo++//z7nzp3DaDRiNBoZNmyY1cfo06cPK1euZM2aNVSpUoUOHTqk9MEKCwtj5syZT/QKE/mAhi0iRD73eB+spyGNPliqqqr3799XP/jgA7Vs2bKqvb296uHhobZt21bdu3dvmvcTFhamDh48WPX19VUdHR3VGjVqqCtXrrTac0hVVfXcuXPqwIED1eLFi6v29vaqp6en+sILL6gjR45UDx48+ETOjPQoioiIUKdMmaI2atRILVSokGpvb68WLlxYbdOmjbp58+Y0b7N161a1Y8eOqq+vr2pnZ6cWLFhQrVu3rjpt2jT12rVrKdul10MoPDxcnThxolq5cmXVyclJdXV1VcuUKaP26tVLXbly5RPb3759W33rrbfUkiVLqvb29qqXl5dau3Zt9csvv0y13fnz59V27dqpHh4eqqIoqXoJpffa7tmzR23btq3q4eGh2tvbq2XLllU/+OAD9f79+09sm95rm16Pqccl98G6cuWK1W0y+jodOHBAHTJkiFq5cmXVw8NDdXJyUsuUKaOOHDky1XuSbPv27Wrt2rVVBwcH1dvbW+3Tp496+/btTPXBunXrlvrSSy+pPj4+qk6ns/raZpXr16+rb731llqmTBnVwcFBdXd3VytUqKC+/vrr6vbt21Nta+09Sq9XmrX91WQyqZMmTVKLFSuWsm/Mnj1bvXz5cob7YFksFnXevHlqpUqVVEdHR7VgwYLqwIED1Tt37qS5vaqq6sKFC9UXXnhBdXBwUIFU+5W12yQlJakzZsxIuZ2bm5vauHFjdc2aNU9sm97/h/T+74rcRVHVZ5j5KIQQQgghrJI5WEIIIYQQWUwKLCGEEEKILCYFlhBCCCFEFpMCSwghhBAii0mBJYQQQgiRxaTAEkIIIYTIYnmqwIqLi+Po0aPExcVpHUUIIYQQ+VieKrDOnTtHjRo1nrmrcHZfbFWIrCL7qshNZH8VuUVW7qt5qsB6Xra4eKoQWUH2VZGbyP4qcous3FelwBJCCCGEyGJSYAkhhBBCZDEpsIQQQgghsphB6wBCCCGEeHbXrl0jNDRU6xh5QlRUFEaj8anb+fj44O/vn+42UmAJIYQQudS1a9eoUKGCtCeyMWdnZ86ePZtukWWTAishIYGXX36ZM2fO4OzsTMGCBZk3bx4BAQEAXLhwgb59+xIaGoqHhwcLFy6kYsWKT10nhBBC5GehoaHExcWxePFiKlSooHWcfOHs2bP07t2b0NBQ7QssgMGDB9O2bVsURWHu3LkMHjyYLVu2ADBkyBAGDx5Mv379WL58OQMHDmT//v1PXSeEEEIIqFChAoGBgVrHEI+wySR3R0dH2rVrh6IoANSpU4fLly8DcPfuXY4ePUrv3r0B6N69O1euXOHq1avprhNCCCGEyKk0mYM1Z84cOnbsCEBISAiFCxfGYHgYRVEU/P39uXbtGi4uLlbXJR9eTMvw4cMxGo1069aN7t27ZzhXRETEsz8pIWxI9lWRm8j+mn2kS752oqKiCA8PT/nZy8sr1XqbF1jTp0/nwoULzJs3L2VZ8shWMlVVM7TOmrlz5z7zUOnjL5AQOZXsqyI3kf01e2TkjDeRPYxGY7r7tU0LrBkzZrBy5Uq2bduGs7MzAMWKFeP69euYTCYMBgOqqhISEoK/vz/Ozs5W1wkhhBBC5FQ2azT65ZdfsnTpUrZu3YqHh0fKcj8/P6pXr87ixYsBWLFiBQEBAQQEBKS7TgghhBAip7LJCNb169cZO3YsJUuWpGnTpgA4ODgQHBwMwPz58+nXrx/Tp0/H3d2dn3/+OeW26a0TQgghhMiJbFJgFS1aNN25U+XKlbPaeiG9dUIIIYTI+QICAoiLi+PGjRvY2dkBsGPHDpo3b87YsWOZMWNGpu6vX79+1KxZk+HDh6e7XZMmTRg3bhwdOnR4Yl2PHj3Yt28ft27dIiYmBldX10xleBq5FqEQQgghsp2/vz9r165N+XnBggXUrFlTszxDhw7l+PHj2Xb/UmAJIYQQItsNGDCABQsWAA9bHBw4cIA2bdqkrDebzYwbN47KlStTuXJlRowYQWJiIgA3btygefPmVKlShc6dO6e69mJMTAyDBg0iKCiIKlWqMHToUJKSkp6ap0WLFvj5+WXxs/wfuRahEEIIkcfdilO5lQ2XKyzkDIWcladvCDRq1Iivv/6aGzdusG7dOl588UX0en3K+u+++44jR45w5MgR9Ho9nTp1Yvbs2YwfP56RI0fSqFEjJk+ezOXLl6latWpKcTZ27FgaNWrE999/j6qqDBo0iLlz5zJ69Oisf8KZIAWWEEIIkcfNP2th6lFLlt/v5EAdU2ron77hf/Xp04eff/6Z1atXs2TJEpYsWZKybtu2bQwcOBAHBwcABg0axLx58xg/fjw7d+5kzpw5AJQsWZLmzZun3G716tUcOHCAmTNnAhAfH4+9vX1WPL3nIgWWEEIIkccNqaCjU/GsnxVUyDlz2/fr14/AwEDKli1LmTJlUq1TVfWJ5uKP/5wWVVVZvXo1JUuWzFyYbCYFlhBCCJHHFXJWMl0MZYfChQvzySefUL58+SfWtWzZkoULF/Liiy+i0+n48ccfadGiBQDNmjVjwYIFfPDBB1y9epXt27enrOvUqROffvop3377LQaDgYiICMLCwihdurRNn9vjZJK7EEIIIWymf//+1K1b94nlgwcPpmrVqgQGBlKtWjUCAgIYOXIkALNnz2bXrl1UqVKFcePGpRRXALNmzcJgMFCtWjWqVKlCixYtuHr16lNzdOrUiaJFiwIPW0I1adIkS55fMkXNyMX9comjR49So0YNjhw58kzXIgwPD5frZYlcQfZVkROoFjOW6AhMkaGYo0IxR4Zijrz38O///qw+iMf+1Qn4VHy268OK9D3v7z2ReRl9zeUQoRBCiCeoZhPmqLD/Fk1WCqjocLA8MnHazh6Dhy96Dx8MXgVxKFmZ+L//JHHTItQK1TM0n0aIvEIKLCGEyGdUUyLmyLBHiqaHhZPpkZ8tMRHwyAEOxd4RfXLxVKAYDuWqozf6oPf87zKjL4qz6xNFlEOpFwidP5H4E3/hXLWBrZ+qEJqRAksIIfIQS2LCI0XTYyNP//235X5UqtsoTq7oPXzQG32wL1wSfcWglJ/1Hj7oPXxRHJ2faQTKsUJNdGWqE7Xme5wqBqHYaX/6vBC2IAWWEELkEpaEOOuH65JHnuJiUt1G5+KeUijZ+5dDX6V+ykhU8nKdg1O25rZv+SoJ894hZtdK3Fu+nK2PJUROIQWWEEJoTFVV1Pj7mCND058wnpC6FbfOzfN/I08lK/93tOnhzwYPX/RGbxR7B42e1SM5fQrj2rATMVv/g0tQS/RGb60jCZHtpMASQohspKoqltjop448qYkJ/7uRoqBz9/pvoeSDXdnq/xt1Sh55MnqhGHLP4Tb31r2IO7ydqA0/49VrjNZxhMh2UmAJIcQzUi0WLPcj05zv9OhIFKZHLjyr0/1vbpPRB7vCJVMXTh6+6N09UfR56+NZ5+yGe9vXiFzxDa4NO2JfrMzTbyTyjICAAOLi4rhx4wZ2dnYA7Nixg+bNmzN27FhmzJiRqfvr168fNWvWZPjw4elu16RJE8aNG0eHDh1SLb958yb9+/fn6tWrODg4UL58eebNm5el7W/y1v9gIYTIIhnp8WSOCgOz6X830tuh9/BOObvOvnh5DMnF038LKJ2bB4ou49duy0tc6rbl/l/riVz5f/iOnCltG/IZf39/1q5dS/fu3QFYsGABNWvW1CSLXq/ngw8+oEGDh2e2jh8/ngkTJvDdd99l2WNIgSWEyHdUswlzdDjmiMcP1Vnv8aTYOaQUSsk9nh4fedK5uKPo5AIZ1ih6PR5dBhP6f+8Rf2wPzoGNtY4kbGjAgAEsWLCA7t27ExUVxYEDB3jllVeIj48HwGw2884777Bp0yYAmjZtysyZM7G3t+fGjRu89tpr3Lt3jxIlSmA2m1PuNyYmhjFjxvD333+TkJBAvXr1+Prrr1NGytJSoEABChQokPJz7dq1mTdvXpY+XymwhBB5ynP1ePL0xVDAH4dygf8rntLp8SQyz7FcII6V6xC17kecKtfJEZPw8wNzVNjDLw1ZTO/uleGTFho1asTXX3/NjRs3WLduHS+++CJ6/f9Gc7/77juOHDnCkSNH0Ov1dOrUidmzZzN+/HhGjhxJo0aNmDx5MpcvX6Zq1aq0adMGgLFjx9KoUSO+//57VFVl0KBBzJ07l9GjR2col9ls5ptvvqFLly6Zfv7pkQJLCJFrWBITHv6iSGvkSYMeT+LZGDsP4s6nQ4jZtQL3Vr20jpMv3N/3BzGbl2T5/bq1fhVj2z4Z3r5Pnz78/PPPrF69miVLlrBkyf8ybdu2jYEDB+Lg8LDoHjRoEPPmzWP8+PHs3LmTOXPmAFCyZEmaN2+ecrvVq1dz4MABZs6cCUB8fDz29hk7AURVVd544w08PDwYMWJEhp9HRkiBJYTIEZ6rx5OnL/b+5dFX8XnssF3293gSmWfnWwTXRp2J2fYbLkGt0Hv4aB0pz3Ot1w6nynWy/H717pmbFN6vXz8CAwMpW7YsZcqkPtFBVdUnvuhk5IuPqqqsXr2akiVLZioLwMiRIwkJCWH16tXosvjwvhRYQohs9bDHUyzmqFBMEffyXI8n8WzcW/Ui7tA2otb/hFfv8VrHyfP0Ru8c0X+scOHCfPLJJ5QvX/6JdS1btmThwoW8+OKL6HQ6fvzxR1q0aAFAs2bNWLBgAR988AFXr15l+/btKes6derEp59+yrfffovBYCAiIoKwsDBKly6dbpaRI0dy8eJFVq9eneERr8yQAksI8cyerceT7uG8jf8WTHYFUs93etjjyRvFYH2Cqsj9dE4uuLfrS+SyObg27IR98XJaRxI20r9//zSXDx48mEuXLhEYGAg8bLEwcuRIAGbPns1rr73G77//TtmyZVOKK4BZs2bxzjvvUK1aNXQ6HXZ2dnz22WfpFlh//fUXX3/9NeXLl6d27doAlChRglWrVmXV00RR1UdmeuZyR48epUaNGhw5ciTlDcqM8PDwLO2BIUR2scW+qlosWGKjHhZJj408We/xpH/4TTnVHKe83+NJpM/a/qpazNydMQLF3gHft76UeXDP4Hl/74nMy+hrLp9yQuRDz97j6X8Fk/R4Es9L0ekxdh1C6DfvEH90F841mmodSYgsIwWWEHlMSo+nyMcKp+ft8eRqlBEGkeUcy1TFqUp9otb9iGPluugcHLWOJESWkAJLiFxGtZgxXztP3JWEjPV4cnBKuY6d9HgSOZGx0+vc/mQwMTt+z9Qp/0LkZFJgCZGLWGJjCFv0CQ/OH+UBT+nx5PmwqNI5umgdW4h0GXwK4dakK/d3LMelThsMnr5aRxLiuUmBJUQukXTzCqE/foiaEIvDK+PxrlZPejyJPMOt5cvEHtxK1PoFePd5R+s4Qjw3uWiWELlA3PG93J01Cp2jM35j5qAvU12KK5Gn6BydMbbvR/yRnTy4ckbrOEI8NymwhMjBVIuZqPU/Eb7wYxwr18X3rZkYvAtqHUuIbOEc1AK7oqWIXDUP9ZGTMITIjaTAEiKHssTdJ+z7KcRs/x1jp9fx6vMOOns5w0rkXYpOj0fXYSRd+4e4Izu0jiPEc5E5WELkQEm3/yXsxw8xx0bhM2QajuVraB1JCJtwKFUZp2oNiVr/E05V6suhcJFrSYElRA4Tf2If4Uu+wOBVgAJj5mDwKax1JCFsythxILc/GUTM9mUY2/XVOk6ucPbsWa0j5BsZfa2lwBIih1AtFqI3LyFm8xKcqjbAs9dY+fYu8iWDd0HcmnYnZueKh20bvApoHSnH8vHxwdnZmd69e2sdJV9xdnbGx8cn3W2kwBIiB7DExxK++HMSzhzEvX0/3Fr0lMafIl9za9GT2OAtRK1bgHffd7WOk2P5+/tz9uxZQkNDtY6S7b7428yaf1VWt9Tj45Q9n49RUVEYjcanbufj44O/v3+620iBJYTGku6EEPbjVMzREXgP+hCnirW0jiSE5nQOThg79Cfi15k8aNgRh5KVtY6UY/n7+z/1l31udzZC5fejJj7qoqNVtey73qm1C5M/CzmLUAgNxZ86wN2v3gJFh9/YOVJcCfEI55rNsStWhshV86VtQz43NtiMvyuMqpx7ypZMJb169Srbtm0jPDw8u/IIkS8kz7cK+2EKDmWq4Td6Fna+RbSOJUSOouh0eHQdSlLIBeIObdM6jtDIxhALG0NUZtTW42jIPVMnrBZYY8eOZdSoUSk/r1q1inLlytGqVSvKlCnDkSNHbJFPiDzHkhBH2MKPiN74C+5t++DdfyI6R2etYwmRIzmUrIRT9cZEbfgJS0Kc1nGEjSVZVMYcMNOkkELXgNxTXEE6BdaqVauoWbNmys/vvfce7dq148SJEwQFBTFx4kSbBBQiLzHdu8ndWaN4cP443q9Pxr31qyi63DPkLYQWjJ0GosbHEbPtN62jCBv7vzMWzkfCV3X1ue7EH6uf7Ldu3UqZNHfp0iXOnz/PxIkTqVy5MiNHjuTw4cM2CylEXpBw9jB3vhwJZjN+Y2bjVLmu1pGEyBUMnn64NutBzK6VmMJuax1H2EhYgsqUoxZeL69QzTt3FVeQToFlNBq5e/cuAFu3bsXLy4saNR52k7a3tyc+Pt42CYXI5VRVJXrbb4R+9wEOJSvhN2YOdgWKaR1LiFzFrdmL6F2MRK39QesowkYmH7FgtsBHNbPvrMHsZLVNQ6NGjZg0aRJ37txhxowZdOnSJWXd+fPn8/wpoUJkBcuDBCKWfkn88T24teqFe5veckhQiGegc3DEvWN/IhZ/wYOLJ3AoXUXrSCIbnQ5XmXfWwqdBOvyyqedVdrP6Sf/VV19RsGBBJkyYgL+/Px9//HHKul9++YWGDRvaJKAQuZUp9Bb3Zo0i4ewhvPtPxNjuNSmuhHgOzoFNsS9e/r9tG8xaxxHZRFVVRh8wE+AGIyrl3s9MqyNYRYoUYceOtK9mvnnzZhwdHbMtlBC5XcL5o4T//Ak6Zzf8Rn2FXaEArSMJkespOh3GrkO4N2s0ccFbcKnbVutIIhtsuKay9cbDju0O+tw5egUZ7IMVEhLCvn37iI2NBcDd3R17e/tsDSZEbqSqKjE7lhM6byJ2/uXwGzNbiishspBDQAWcazQlasPPWBJitY4jslii+WFbhuaFFToVz73FFTylwPruu+8oUqQIxYsXp2HDhpw/fx6Arl27Mnv2bJsEFCK3sCQmEL74c6LW/oBbsx74DJ6KztlN61hC5DnuHQegJsYTvWWp1lFEFpt72sKlmNzZluFxVgusWbNmMWLECHr37s3GjRtRVTVlXZMmTVi2bJlNAgqRG5jC73Bv9lgSTu7Dq+97GDsOQNHlzjNfhMjpDB6+uDV7ifu7V2O6d1PrOCKL3ItX+fCYhcHldbzglbuLK0inwPr666/54IMP+Oyzz2jRokWqdeXKlUsZzRIiv0u48Dd3Z47EkhCL76ivcK7eSOtIQuR5rs26o3fzJFLaNuQZk448vN7khzVy78T2R1l9Fjdu3KBevXpprrOzs0uZjyVEfqWqKjG7VxP6f+9iV6QkfmPmYF+4pNaxhMgXdPaOGDsNJOHkPhL+Oa51HPGcToSpfHfOwuRAHb65tC3D46wWWMWLF+fgwYNprgsODqZs2bLZFkqInE5NfEDErzOJWjUP18Zd8RnyEXoXd61jCZGvOFVvjH2JikStmodqlrYNuVVyW4bS7vBmxbwxegXpFFiDBg3io48+4scffyQ6OhqApKQkNmzYwBdffMGQIUNsFlKInMQUcY+7X48j7vgevPq8g0fnQSh6mW8lhK0pioJH1yEk3bpK7IFNWscRz2jNvyo7bqp8WUePfS5uy/A4q32wxo0bx7Vr1xg8eHBKMVW/fn0A3njjDd544w3bJBQiB3lw6RRhP32EYmeP38iZ2Bcro3UkIfI1e/9yONdqQfQfi3Cu3hids6vWkUQmPDCrjAs206qIQrtieae4gnQKLIA5c+YwatQotm7dSlhYGF5eXjRv3pwyZeSXishfVFUl9q/1RK6ch0PJSnj1ew+9q4fWsYQQgLFDf+L//pPoLb/i0WWw1nFEJsw5ZeFqDKxtlfvbMjwu3QILoGTJknI4UORrqimRiOXfEHdgM66NumDs/DqK/qn/dYQQNqI3euPWoifRmxbjUq8ddn5FtY4kMuBOnMq0YxaGVdBR0TNvFVfwWIF19OjRTN04MDAwS8MIkdOYo8IIWzCNxBuX8Ow1FpegllpHEkKkwa1JN2L3byRqzff4DJqqdRyRARMPmzHoYEoeacvwuFQFVs2aNTM0RKeqKoqiYJazNkQe9uDKGcJ+moai0+M3cgb2/uW0jiSEsEKxd8DY+XXCF04n4fxRHMvJAEBOdjxM5cfzKrPq6vB2zHujV/BYgbVz506tcgiRo9zfv5HI5d9gX7wc3v0nonfz1DqSEOIpnKo2xL5kZSJXzafA+G/l7N4cSlVVRu03U94DhuWhtgyPS1VgNW7cWKscQuQIqimJyJX/R+y+P3Bp0AGPLkNQDHZaxxJCZEBy24a7X44kdt8GXBt20jqSSMPKqyq7b6lsbKPHTpc3R68gnT5YISEhVudkHT16lOvXr2dbKCG0YI4O59437xAbvBXPnm/h2WO4FFdC5DL2xcrgHNSS6I2/YImL0TqOeEyCSWXcATPtiim0KZZ3R68gnQJr2LBh/PLLL2mu+/XXX3nzzTezLZQQtpb473nuzhyJKew2viM+x6VuW60jCSGekbF9P1SzmehNS7SOIh4z65SF67Ews07eP3xrtcAKDg6mWbNmaa5r2rQp+/fvz7ZQQthSbPAW7s4Zh97TlwJjv8YhoILWkYQQz0Hv7oVby5e5/+c6ku6EaB1H/NetOJWPj1t4s5KO8h5599BgMqsF1v3797GzS/vwiE6nIyZGhl5F7qaaTUSs+JaIpV/iEtQC3+GfoTd6ax1LCJEF3Bp3Qe/pS9Tq77SOIv7r/UNmHHQwOTBvHxpMZvVZVqhQgVWrVqW5bvXq1ZQrJ6esi9zLHBPJvW/fJXbfH3i8OALPnm+hGOy1jiWEyCKKnT0enV4n4ewh4s8c0jpOvnfknsrCf1Q+rKnD0yHvj15BOp3cR40aRb9+/dDr9QwYMIDChQtz8+ZNfvrpJ3744QcWLFhgy5xCZJnEkAuE/fghqtmE75uf4lCystaRhBDZwLFKfRxKVyFq9Xc4lqsuV2DQSHJbhoqeMLh8/hi9gnQKrNdee407d+4wdepU5s+fn7LcycmJTz/9lL59+9okoBBZKfbwdiJ+m41doRJ4D5iIwcNX60hCiGyiKArGrkO5O+NN7v+5HrfGXbSOlC/9flnlzzsqW9rqMeThtgyPS7ecHz9+PEOGDGH//v2EhYXh7e1N3bp1cXd3t1U+IbKEajYTtfYH7u9ehXNQSzxfHIFiJ4cEhcjr7IuUxKVOG6I3L8a5ZjP0LvL7y5biTSrjD5rp6K/Qsmj+Gb2CDFzs2d3dndatW9siixDZwnw/ivCfP+HBpRN4dH8DlwYd89xV24UQ1rm360vcsd1Eb/wFzx7SYsiWZp6wcCsOtrXL+20ZHpeqwFq5ciXNmjXDw8ODlStXPvXG3bp1y7ZgQmSFxOuXCPtxKmriA3zf+BSH0lW0jiSEsDG9mwfurXoRtX4BrvXbY1coQOtI+cLNWJVP/rYwopKOMsb896U2VYHVo0cPDhw4QFBQED169Ej3hpm92PPIkSNZu3Yt//77LydPnqRy5f9NLA4ICMDR0RFHR0cA3n33XXr27AnAhQsX6Nu3L6GhoXh4eLBw4UIqVqyY4ccV+Vfc0V1ELP0KQ4FieI+chMHTT+tIQgiNuDbqTOy+P4hc/R0+Qz+WUWwbePeQGWcDfFA9fx0aTJaqwLpy5QqFChVK+XdW6tGjB2+//TYNGjRIc/3y5ctTFV3JhgwZwuDBg+nXrx/Lly9n4MCB0uRUpEu1mIlav5D7O37HuWYzPF96C8XeQetYQggNKQY7jJ0HEfbjVBLOHMSpUm2tI+VpB+9aWHRBZV4DHR75pC3D41IVWMWLF0/z31mhUaNGmb7N3bt3OXr0KFu2bAGge/fuDB8+nKtXrxIQEJCl+UTeYImNIWzRJzy4cBxjl8G4Nu4q31SFEAA4Vq6DQ9lq/23bECjXGs0mD9syWKjiBa+Xy5+jV5DOJHe9Xs/+/fsJCgp6Yt2RI0cICgrK1CHCp3n11VexWCzUrl2bTz75BF9fX0JCQihcuDAGw8OYiqLg7+/PtWvX0i2whg8fjtFopFu3bnTv3j3DGSIiIp73aQgNWe5c48GyL1ET4nDoNYGkkpXz7HuaV5+XyJty0v6qNH0Z03fvcnfLb9jVaad1nDxpRYiB/XedWN0wjqjIrKsTbOF59lUvL69UP1stsFRVtXonSUlJ6PVZd0bAnj178Pf3JykpiYkTJ9K3b1/++OMPgCdGH9LLlWzu3LkEBgY+U5bHXyCRO8Qd30vErzMx+BTGe/jnGLwLah0p28m+KnKTHLO/enkRUa89cXtW4dOoA3pXD60T5SlxJpUPT5voUlyhc3mj1nGeSVbtq6kKrNu3b3Pz5s2Un8+fP58yepQsISGBBQsWZOkhRH9/fwDs7OwYNWoUZcuWBaBYsWJcv34dk8mEwWBAVVVCQkJSthdCtZiJ3vgLMVv/g1P1xni+PBqdg6PWsYQQOZh7297EHdlJ9B+/4PnSCK3j5Clf/G3hbjzMqJP/2jI8LlX1NH/+fKZOnYqiKCiKQr9+/Z64gaqq6PV6vv322ywJEBsbS1JSEh4eHgAsXbqU6tWrA+Dn50f16tVZvHgx/fr1Y8WKFQQEBMj8KwGAJe4+4Ys/I+HsEYwdB+LarIfMtxJCPJXe1QP3Nq8SteYHXBq0x75wSa0j5Qkh91U++9vCqMo6SrnLZ3GqAqtfv340adIEVVVp1qwZ33zzzRMtEezt7Slbtize3t6ZeqA333yTNWvWcPv2bVq0aIGrqysXL17kzp07dO/eHbPZjKqqlCxZkkWLFqXcbv78+fTr14/p06fj7u7Ozz///BxPV+QVSbevEfbjVMz3o/AZ/CGOFWpqHUkIkYu4NuhI7F8biFr9HT7DPpEvZ1lgwkEzbvbwfj5ty/A4RU1jUlNiYiJr166lXLlyvPDCC1rkeiZHjx6lRo0aHDly5JnmYIWHh+eceQLCqviT+wlf/AV6T198Bk7G4FtY60g2J/uqyE1y6v4afzqYsO8n4z1wMk4v1NU6Tq62/46FemvNfN9Qz+u5+ILOWbmvpvkq2Nvb07t3b8LCwrLkQYTICqrFQtTGXwj7cSqO5QPxGz0rXxZXQois4VgxCIfyNYhc8x2qKVHrOLmW5b9tGap5Q/+yMhKYzGqZWb58eUJCQmyZRQirLAmxhC34kJgtv+Levh9e/d5H5+CkdSwhRC6mKAoeXQZjDr/D/T1rtI6Tay25qHLwnsrsunr0OimwklktsD755BM++ugjjhw5Yss8Qjwh6U4Id78cxYOLJ/F+fQruLV+W+RJCiCxhV7A4LvU7EL15KeaYnNOvK7e4n6Qy4aCZHiUUGhXKvYcGs4PVPlhvv/02oaGhBAUF4ePjg5+fX6pfaoqi8Pfff9skpMi/4k8HE/7LZ+iNPviNmY2dX1GtIwkh8hj3Nr2JO7KD6D8W4dnzLa3j5Cqf/20h7AF8HiRtGR5ntcCqUaMGNWvKmVlCG6rFQsy2/xC98RccK9XBq/c4dI4uWscSQuRBehd3jG36ELlqHi71O2BftJTWkXKFf2NUvjhhYcwLOkpIW4YnWC2wFi5caMMYQvyPJSGOiF9nEn/iL9zb9MatVS8UnQw9CyGyj0v99tz/awORq+bhO/xzmYaQAe8cNONhD+9Wlc/ntMirInIU072b3J01moTzx/AeOBn3Nr2luBJCZDtFb8Cjy2ASL50k4cRfWsfJ8f68beG3yyqfBOlxs5diNC1WR7AAIiMjWb58Of/88w8JCQlPrJ8zZ062BRP5T8LZw4Qt+hS9qxG/0bOwKyiXRBJC2I5jhZo4VqxF5NofcKwYhGJnr3WkHCm5LUMNH4XXykhxZY3VAuvChQvUq1ePBw8eEBsbi6+vL+Hh4ZhMJjw9PTEajVJgiSyhqiox238nesNCHCvUxKv32+icXbWOJYTIh4ydB3Pns6HE7F6Fe4ueWsfJkRZdUDkSqrK3ox6dHEq1yuqxlzFjxlC7dm3u3LmDqqr88ccfxMfHs3jxYtzc3Pj9999tmVPkUZYHCYT//AnR6xfg1uIlvF+fIsWVEEIzdgWK4dqgIzFb/4M5SpptPy4mUeXdg2Z6llRoUFCmb6TH6qtz8OBBhg4dioODA/Dw8jl6vZ5evXoxduxYRo4cabOQIm8yhd3m3uzRJJw5iFf/iRjb95P5VkIIzbm3eRXFYEfUBrn27eM++dtCZCJ8Jm0Znsrqb7MHDx7g7u6OTqfDy8uLmzdvpqyrVKkSx48ft0U+kUcl/HOMuzNHYHmQgN/oWThXbaB1JCGEAEDn7IZ72z7EHdpKYsgFrePkGFeiVb48aWFcFR3F3eTQ4NNYLbDKli3Lv//+C0D16tX59ttviYmJIT4+nvnz51O4sFwDTmSeqqrE7FxB6P+9j12xMhQYMwe7QgFaxxJCiFRc6rbDUMCfyFXzUFVV6zg5wtsHzXg7wDvSliFDrL5KL7/8csoo1bRp0zh06BCenp64u7uzYsUKJk+ebKuMIo+wJCYQsfhzotZ8j1uzHvgMmYbOxU3rWEII8QRFr8ej6xASL58m/vhereNobvctC8uvqHwapMfVTkavMsLqWYRjxoxJ+XedOnU4deoUGzduJCEhgWbNmlG5cmWbBBR5gyn8DmELPsR05zper72Lc2BjrSMJIUS6HMsF4lipNlFrf8CpUm0UewetI2nCbFEZtd9MkK/Cq6WluMqodPtgPapYsWIMHjw4O7OIPCrhwt+EL5yO4uCI76ivsC9SUutIQgiRIcYug7nz6RBidq3EvdUrWsfRxMJ/VI6Hwf5OOmnLkAnpFlhJSUn8/PPPHDhwgFu3blGoUCHq1KnDa6+9hr29NGAT6VNVldi9a4lcPR+H0lXw6vseehd3rWMJIUSG2fkWwbVRZ2K2/YZL7Vbojd5aR7Kp6ESV9w6b6VVKoU4BmXuVGVZfrX/++Ydy5coxdOhQDh8+TFJSEocPH2bo0KGUK1eO8+fP2zKnyGXUpEQifp1J5Mr/w7VRF3yGfCzFlRAiV3Jv+QqKnQNR63/SOorNfXzMQkwifCptGTLNaoE1ZMgQ7O3tOXfuHMePH2fLli0cP36cs2fP4ujoyLBhw2yZU+Qipsh73P16HHHH9+DZezweXQaj6OU/pxAid9I5u+Le7jXiDm0j8d/8M7hwKVpl1ikL71TVUcxVDg1mltUCKzg4mI8//pjSpUunWl6mTBk+/PBDDhw4kO3hRO7z4PIp7s4YgSU6Ar+RM3Gp2VzrSEII8dxc6rbBrnCJfNW2YXywGT8nGC9tGZ6J1VetcOHCKFYms+l0OgoWLJhtoUTuo6oq9//awL2572AoUAy/sV9jX6yM1rGEECJLKDo9xq5DSbx6lviju7SOk+123rSw6qrKZ0F6nA0yevUsrBZYkydP5oMPPuDy5cupll+6dIlJkyZJHyyRQjUlEvnbbCJ//xrX+u3xfeMT9G4eWscSQogs5VimKo4v1CNq3Y9YEhO0jpNtktsy1PVTeKWUFFfPyupZhMuWLSMyMpJy5cpRuXJl/Pz8uHv3LqdOnaJAgQKsWLGCFStWAKAoCmvWrLFZaJFzmKPCCFswjcQbl/B8ZQwutVtpHUkIIbKNR+fXuf3JEO7vWI57m95ax8kWP5y3cCIcDnbRWT2SJZ7OaoF1//59ypYtS9myZYGHF3v28PCgQYOH14yLiYmxTUKRYz24coawnz5CUXT4jZiBffFyWkcSQohsZfApjGvjLsRs/x3nOq0xePhqHSlLRSWqTDxs4bUyCrV8Ze7V87BaYO3cudOWOUQuE7t/IxHLv8Hevyze/Seid/fSOpIQQtiEe6uXiTu4leh1C/Dq847WcbLUtKMW4kwwvZac+f28pDwVmaKakoj4fS4Rv83GpXZrfN/8TIorIUS+onN0wdi+L3FHdvLg6lmt42SZC1Eqc05beLeqjiIucmjweUmBJTLMHB3OvW8nEHtgEx4938LzpREoBjutYwkhhM05126FXZFSRK6ch2qxaB0nS4w9YKawM4ytIqVBVpBXUWRI4rXz3J05ElPoTXyHf45r3bZaRxJCCM0oOj0e3YaSdO08cUd2aB3nuW29bmHdNZXPg/Q4SVuGLCEFlniq2OAt3J0zDr2HDwXGzsWhREWtIwkhhOYcSr2AU9UGRK3/CcuD3Nu2wWRRGX3ATIMCCi+WlOIqq0iBJaxSzSYiV/4fEUu/xLlmM3xHfJ7vLnQqhBDpMXZ6HUtsNDHbl2kd5Zl9d87CmQiYVVcvbRmykNWzCEX+Zr4fSdhPH5N45QwePYbjUr+9/McTQojHGLwL4takGzE7l+NSpzUGrwJaR8qUiAcqkw5b6FdWoYavfMZnpVQF1oABAzJ14wULFmRpGJEzJIZcIGzBNNSkRHzf/AyHUpW1jiSEEDmWW4uexB7cQtS6BXj3fVfrOJky9aiFBxZpy5AdUhVYhw4dSrXy9u3bhIWF4e7untLJPTo6Gm9vbwoVKmTToMI24g7vIPy3WdgVLI73gEkYPPNWEz0hhMhqOkdnjO37E7H0Sx407IRDyUpaR8qQc5Eq35y2MK2mjoLOMnqV1VLNwTp58mTKny+++AJXV1c2b95MZGQk//zzD5GRkWzatAlXV1c+++wzrTKLbKCazUSu/o7wxZ/jXK0RfiNnSnElhBAZ5FyrBXbFyhC5Kve0bRh7wEwxVxhVWaZjZwerr+rbb7/Nhx9+SMuWLVMtb9WqFVOmTGH8+PHZHk7Yhjk2mtD573N/z2qMXYfi2Wssip291rGEECLXUHQ6PLoOISnkAnGHtmkd56k2hVj4I0Tli9p6HKUtQ7awOsn9woULeHml3aHby8uLS5cuZVsoYTuJNy4T9uNU1AcJ+Az7BMcyVbWOJIQQuZJDyco4VW9M1IafcKraAJ2js9aR0pT037YMjQspdAuQ4iq7WB3BqlixIp9++ukTF3WOiYnh008/pWJF6YWU28Ud3c29WaPRObviN3aOFFdCCPGcjJ0GYomPzdFtG+adsXA+UtoyZDerI1hff/01bdq0oVixYjRt2jRlkvvOnTsxm81s2rTJljlFFlItZqI3/EzM9mU41WiKZ8+30Nk7ah1LCCFyPYOnH25NexCz43dc6rTB4F1Q60iphCWoTD5qYWA5hWreUlxlJ6sjWPXq1ePChQsMHTqUqKgo9uzZQ1RUFEOHDuXChQvUr1/fljlFFrHExRD63SRidizH2HkQXr3fluJKCCGykFvzF9G5uBO19getozxhyhELJgt8VFPaMmS3dBuNFihQgE8//dRWWUQ2S7p1ldAfp6LG3cdn6Ec4lgvUOpIQQuQ5OgcnjB36E7FkBg8unsChdBWtIwFwOlzl/85a+KSWjgLSliHbPfXczIiICPbu3cuvv/5KREQEAAkJCVhyyWmo4qH4v//k7lej0Nk54DdmjhRXQgiRjZxrNMPOvxyRq+ejWsxax0FVVcYcMBPgBiOlLYNNWH2VLRYL7733HsWKFaNx48b06dOHK1euANCtWzemTZtms5Di2akWC1EbFhL200c4VgzCd9QsDD7SJFYIIbKTotPh0W0oSdcvEXdwq9Zx+CNEZcsNlRm19TjoZfTKFqwWWJMmTWLu3Ll88cUXnDlzBlVVU9Z16tSJdevW2SSgeHaW+FjCfphCzLbfcO8wAK++76JzkPlWQghhCw4BFXCq0ZSo9QuxJMRqliPR/HD0qllhhc7FpbiyFasF1sKFC5k+fTrDhg2jTJkyqdaVKlVK+mDlcEm3r3H3y5E8uHIGn8HTcG/xkpyOK4QQNmbsMAD1QTzRW/6jWYZvzli4GC1tGWzNaoEVFhZGhQoV0lxnsVhISkrKtlDi+cSf2s/dr0aB3kCBMXNwrFBT60hCCJEvGTx9cWv+Ivd3r8YUetPmjx+aoDL1qIXB5XW84CXFlS1ZLbDKli3L1q1pHzfeuXMnlStXzrZQ4tmoFgvRmxYT9sNUHMtVx2/UVxh8C2sdSwgh8jXXZj3QuxmJ1KBtw6TDD09I+7CGTGy3NattGkaPHs2gQYOws7OjR48eAFy/fp39+/czZ84cFi5caKuMIgMsCbGEL55BwukDuLd7DbcWL6Po5D+UEEJoTWfviLHjQMJ/+YyEC3/b7KoZJ8NV5p+z8EWQDl8nGb2yNasFVr9+/QgPD2fKlClMnz4dgC5duuDs7MxHH33ESy+9ZLOQIn1Jd68T9uOHmKNC8X59Ck6VamsdSQghxCOcAptgv3ctUavm4TBuLoouext9qqrK6P1mSrvD8EryZVsL6TYaHTNmDIMHD2bfvn2Ehobi5eVFvXr1cHd3t1U+8RTxp4MJ/+Vz9O6e+I2ejV2BYlpHEkII8RhFUfDoNpS7X75F7P5NuNZvn62Pt+6ayvabKuta6bGXtgyasFpgLVq0iPbt2+Pt7U2rVq1SrQsPD2f9+vW89tpr2R5QpE1VVWK2/Ub0Hz/jWDHo4SVvnFy0jiWEEMIKe/9yONdqQfQfi3AObJJtn9kPzCpjD5hpWUShvb8UV1qxOm7Yv39/q60Yrly5Qv/+/bMtlEif5UE84Qs/JnrDQtxa9cJ74GQproQQIhcwtu+HmpRA9JZfs+0xvj5t4UoMfFVH2jJoyWqB9Whj0cdFRETg5uaWLYFE+kyhN7k7azQJ547gPWASxrZ9ZDK7EELkEnoPH9xa9OT+njUk3buR5fd/N15l2lELQyvoqCRtGTSV6hDhxo0b2bhxY8rPM2fOpECBAqlukJCQwI4dO6hWrZpNAor/STh7mLBfPkXv7I7f6FnYFSyudSQhhBCZ5NakO7H7NxG1+jt8Bk3N0vv+4LAFvQ6mSlsGzaUqsP7555+US+AoisLevXtxcHBIdQN7e3sqV66ccmahyH6qqnJ/53Ki1v2EY/kaePV5B52zq9axhBBCPAPF3gFjp9cJ/3k6CeeP4lguMEvu9+8wlR/OW/iqjg5vRxm90lqqAuutt97irbfeAqBEiRKsXr2aqlVt069DpM3yIIGI/3xF/LHduLV8Gfe2fbL99F4hhBDZy6laQ+z3ViJy1XwKjP8WRf98n+uqqjJqv5myRhhWUUavcgKr78KVK1ekuNKYKew292aPIeH0Abz6vYexfT8proQQIg9QFAWPrkMx3blG7P4/nvv+Vl1V2XVL5cs6eux0MnqVE6TbBys0NJRZs2Zx4MABbt26RaFChahTpw5vvfUWvr6+tsqYLyX8c4zwnz9BcXTBb9Qs7AqX0DqSEEKILGRfrAzOtVoSvfGXh20bnJ/t5LEEk8q4YDNtiym0LSajVzmF1XciODiYMmXKMGfOHFxdXalfvz6urq7MmTOH0qVLExwcbMucNrHsmoHQBOtnT9qCqqrE7FpJ6P+9j13R0hQYM0eKKyGEyKOMHfqhmpKI3vzsbRtmn7Jw7T58WUeOcOQkVkew3nzzTSpVqsSGDRswGo0py6Oiomjbti3Dhw/n0KFDNglpC5EPVCadcODTsyZWtTRQ3cf2Q6xq4gMils0m7vAOXJv1wNi+/3MflxdCCJFz6d29cGv5MtF/LMKlXrtMX43jdpzKR8ctvFlRR3kPOTSYk1gdwTp9+jQTJkxIVVwBGI1GJkyYwKlTp7I9nC15OChsaxaHlwPUX2ti6UWLTR/fFHGXu3PGEv/3X3i9NgGPTq9LcSWEEPmAW+Ou6D19iVrzXaZv+/4hM/Y6mBwohwZzGqvvSOnSpYmMjExzXVRUFCVLlsyuTJop6qzyVycD3Uso9NppZnywGZMl+w8ZPrh4grszR2CJjcb3rS9xDmyS7Y8phBAiZ1Ds7DF2ep2EM4dIOHs4w7c7Gqry0z8qH9bQ4SVtGXIcqwXWF198weTJk9m9e3eq5bt27WLKlCnMmDEj28NpwcmgsKiJni/r6PjqpIV2m8yEZ9O8LFVVub93Lfe+fRe7QgH4jf0a+6KlsuWxhBBC5FxOVepjX+oFIld/h2o2PXX75LYMFTxgSAUZvcqJrL4r48ePJyoqimbNmuHl5UW5cuXw8vKiefPmREVF8c4771ClShWqVKmS59o5KIrC6Bf0bG6r50ioSq3VJk6GZ22RpSYlErH0KyJXfItrw074DJ2O3tX49BsKIYTIc1LaNtwNIfavDU/dfvkVlb23Vb6qq8cgbRlyJKuT3GvUqJHvLxLZvIiOw10Uum41UWeNiYWN9bxY8vm/KZgi7xG24COSbl7G89VxuNRqkQVphRBC5Gb2RUvhUrs1UZt+walGU/Qu7mluF29SGR9spoO/QquiMnqVU1ktsBYuXGjDGDlXCXeFvzoZGLjHzEvbzbwbpjKthg79M35jeHD5NGE/fYSi1+M3cib2/mWzOLEQQojcyr19X+KO7SF602I8u7+R5jZfnrRwMw62tJUToXKyDJW+qqpy8+ZNTKanHxfOi1zsFJY20/NZkI7P/rbQcYuZiAeZP2R4/68N3PvmHQx+RfEbO1eKKyGEEKno3Txxb/0KsX+tJ+n2v0+svxmr8slxCyMq6SgrbRlytHQLrM2bN1OnTh0cHR0pVqwYJ06cAGDw4MEsWbLEJgFzCkVReLuqnj9a69l/RyVotYnTGZyXpZoSifhtNpG/f41LvXb4vvEJejeP7A0shBAiV3Jt1Bm9V4GHE97V1L9n3jtkxskAH1SXQ4M5ndV3aOnSpbRr147ixYszZ86cVG9yqVKl+Omnn2wSMKdpXUzHoS4GHPRQZ62JVVfS75dljgrj3tx3iD24Dc+XR+PZ/Q0UfbpXKBJCCJGPKQZ7PDoN4sG5IyScOZiy/NA9Cz9feDhNxcNBRq9yOqsF1rRp0xg1ahS//fYbr7/+eqp1lSpVynONRjOjtFHhQGcDrYoodNtmZvIRMxb1ydGsB1fPcmfmCEwRd/Ed8QUudVprkFYIIURu4/hCXRzKVCNq9XeopqT/tmWw8IIXvF5eRq9yA6vv0uXLl2nXrl2a61xcXIiKisrUA40cOZKAgAAURXmiOLtw4QL16tWjbNmyBAUFcebMmQyt05KrncLyFno+qqlj2lELXbaYiUr8X5EVe2AT975+G4N3QQqM/RqHgPIaphVCCJGbPGzbMART6C3u/7mO3y6r7LujMquOtGXILawWWAULFuTcuXNprjtx4gTFixfP1AP16NGDP//8M83bDRkyhMGDB/PPP//w9ttvM3DgwAyt05qiKLxfXc+61np231KpvdrE+dBEIpbPJeI/s3Cp3RLfNz9D7+6ldVQhhBC5jF3hErjUbUv0piV8/Gc4nYsrNCsio1e5hdV3qlevXkyZMoXt27enLEseffr888/p3bt3ph6oUaNGFC1a9Inld+/e5ejRoyn31717d65cucLVq1fTXZeTtPfXcbCLAc+kCM5+9S4x+zbh8dJIPF8aiWKw0zqeEEKIXMq9XR8SzPDqv0uYUVvaMuQmVmdbT5kyhdOnT9OyZUu8vb0BaNu2Lffu3aNDhw5MmDAhSwKEhIRQuHBhDIaHURRFwd/fn2vXruHi4mJ1XUBAgNX7HD58OEajkW7dutG9e/cMZ4mIiHjm5+F14xLLzn1FdJKFbsWn00otyZiwcGQkV2SH59lXhbA12V+f3Y04hXk+PXnn1kKcbzYm3OyvdaQ87bnqAK/UR6usFlj29vasWbOGnTt3snXrVkJDQ/Hy8qJFixa0aJG1nccf7xj/6BmL6a2zZu7cuQQGBj5TlsdfoIyIPbiNiGWzcShSirL93qfjZS+mHLVwLs6RnxvrcbOXKktkvWfZV4XQiuyvz+atv03sKNSB95M2o+78D57DPsn3V1nJblm1rz61X0DTpk1p2rRpljxYWooVK8b169cxmUwYDAZUVSUkJAR/f3+cnZ2trssJVLOJqDU/cH/Papxrt8LzxeEoBnsm14Bq3gp9dpmps8bEmlYGShvlP4QQQoiMO3DHwuKLKt83dMSzxmDCvp9MwukDOFWuq3U0kQFW52Bt377daq+rhQsXsnPnziwJ4OfnR/Xq1Vm8eDEAK1asICAggICAgHTXac18P5LQee9z/891eHR/A8+XR6MY7FPWdw7QEdzZQJIFaq02sSkk/X5ZQgghRDKLqjLqgIWqXtC/rIJjxSAcygUStfp7VFOi1vFEBlgtsCZOnMidO3fSXHfv3j0mTpyYqQd68803KVq0KNevX6dFixaULl06Zd38+fOZP38+ZcuW5dNPP+XHH3/M0DqtJF6/yN2ZI0m6dRXfNz/FtWGnNIdsK3gqHOxioH4BhXabzHx63JyhQ5xCCCHyt18vqgTfVZldT49epzxs29BlMKbw29zfs1breCIDFNXKb3x3d3dWrlyZ5nyr7du30717dyIjI7M7X6YcPXqUGjVqcOTIkWeagxUeHv7UY69xR3YS8Z9ZGAr64z1gEgZP36fer0VVmXTYwsfHLbxUUmFBIz0udnLIUDy7jOyrQuQUsr9mTmySSrnfTdTxU1jeIvVMnojl3xB3eDsF318gl1zLBlm5r1odwVIUxWoz0YiICMxmc5YEyC1Us5nINd8T/stnOFVtgN+IGRkqrgB0isJHtfQsb6FnwzWVumtNXI6WkSwhhBBP+vyEhXvx8EXQk20Z3Nv2AUVH9B8/a5BMZIbVAqt27dp88803TxzSUlWVb7/9ltq1a2d7uJzCHBtN6PyJ3N+9CmPXoXi+Og7F3iHT99O9hI4DnQ3EJkHN1Sa2Xpd5WUIIIf7n2n2Vz/+2MOYFHSXcnzzSoXdxx71Nb2IPbCbxxmUNEoqMsnoW4dSpU2natClVqlShX79+FCpUiJs3b7Jo0SL++ecfdu3aZcOY2km8eZmwHz9ETYjDZ+h0HMtWe677q+ylcKiLgVd2mGmzycxnQSpjX9DJabdCCCF456AZD3t4r5r1ju2uDToQ+9cGolbNw+fNz+T3Rw5l9R2sW7cu27dvx93dnXfeeYfevXszYcIEjEYj27dvp06dOrbMqYm4Y3u4N2s0OkcX/MbMee7iKpmXo8IfbfSMr6JjfLCF3jvNxJnkkKEQQuRn++5Y+M8llem10u+fqOgNGLsM5sHFEySc+MuGCUVmpNsHq379+vz111/Ex8cTERGBh4cHzs7OtsqmGdViJnrDz8RsX4ZTYBM8Xx6Fzt4xSx9Dr1P4NEhPNW+FAbvNnIk0sbqlgeJu8k1ECCHyG4uq8tY+C4E+0Lfs038POFWshWOFWkSu/QHHSkGp2gSJnCFDV410cnKicOHC+aK4ssTFEPrdZGJ2LMfY6XW8+ryT5cXVo14upWN/ZwORDx7Oy9p5U+ZlCSFEfvPLBZXDoSqz6ujRZfCQn7HzIMwRd4nZtTp7w4lnYrXAGjhwIC+99FKa61555RUGDx6cbaG0Yrl7nTtfjiTx2jl8hn6EW7MeNjm2XdVb4XBXA1W9FFr+YWb2KemXJYQQ+cX9JJV3D5l5qaRCw0IZGvcAwK6gP64NOhKzdSnm6PBsTCiehdV3csuWLVYvlNytWzc2b96cbaG0YIm7T8LCqejsHCgw5mscyz3btQyflbejwqa2et6qrGPUfgv9dpuJl3lZQgiR531y3ELEA/g8jbYMT+PeujeK3kDUhoVZH0w8F6sF1r179/D1TbvPk7e3t9Uu77mVztkVh65v4jvqKww+hTTJYNApzKyj55cmepZdVmm0zkzIfSmyhBAir7oaozLzpIVxVXTPNAdX5+KGe9s+xB3cSmLIhWxIKJ6V1QKrSJEiBAcHp7kuODiYQoW0KUKyk75MNXQOTlrHoHcZHX91MnAnXqXmahN7b8m8LCGEyIveDjbj5QDvVM34ocHHudRrj6FAMSJXzZfpJTmI1Xf0lVde4eOPP2bZsmWplv/+++9Mnz6dXr16ZXu4/CzQ5+G8rAoeCs02mPn2jMzLEkKIvGTPLQu/X1H5tJYe1+e4fJqi1+PRZQiJl08R//feLEwonofVAmvSpEk0adKEl19+GTc3N8qWLYubmxsvv/wyjRs3ZvLkybbMmS/5OSlsbadnWEUdb/5lYdBeMw/MUmQJIURuZ7aojNpvJshXoXeZ5z+ZyrF8DRwr1SZqzQ+oiQ+yIKF4Xlb7YNnb27N+/Xq2bt3Kjh07CAsLw9vbmxYtWtC8eXNbZszX7HQKc+rpCfRRGPqnmdMRZla00FPYRfplCSFEbvXzBZVjYbCvky7DbRmexth5EHc+G0rMrpW4t3olS+5TPLt0G40CtGzZkpYtW9oii0hHv7I6KnpAt21maqwysaKlnnoFnv2YvRBCCG1EJ6q8d8jMK6UU6mbh57idX1FcG3YiZttvuNRuhd7onWX3LTIvQ+9sXFwc4eHhT/wRthXkp+NwFwOl3BWarDfz/TmZ/C6EELnN9OMWohPhs2doy/A07q16odg5ELX+pyy/b5E5VgssVVX56KOPKFq0KG5ubvj6+j7xR9heQWeFHe31DCynY/BeM0P3mkmUeVlCCJErXIpW+eqkhber6ijmmvVTPXTOrri3e424Q9tIvHY+y+9fZJzVAuurr75i5syZvPnmm6iqyvvvv8+kSZMoW7YsAQEBfP/997bMKR5hr1f4vwZ6vmuoZ8E/FpptMHM7ToosIYTI6d4ONuPnBG8/R1uGp3Gp0wa7QgHStkFjVt/hH3/8kalTp/L2228D0KVLFyZPnszp06epUKECFy9etFlIkbZB5XXs7qDncszDflkH78ohQyGEyKl23bSw8qrKZ0F6nA3Zd6KSotdj7DqUxCtniD+2O9seR6TPaoF19epVqlWrhl6vx87OjsjIyIc30Ol48803WbhwoY0iivTULfBwXlYxF4WG68z8dF6KLCGEyGmS2zLU8VN4pVT2nwXuWLYajpXrErX2RyyJCdn+eOJJVgssb29v7t+/D4C/vz9Hjx5NWXfv3j3i4uKyP53IkMIuCrs66OlTRmHAHjMj/jKTZJFhYSGEyCl+PK/ydzjMrqtDyaK2DE/j0XkQ5pgI7u9YYZPHE6lZbdNQv359Dh06RLt27ejVqxdTpkzh9u3b2NnZ8f3330svrBzGQa/wfUM9NXwsjNxn4US4yu8t9Pg5Sb8sIYTQUlSiysTDZvqUVgjys117HYNvYVwbdyFmxzKc67TC4CEnp9mS1QJrypQp3LhxA4D33nuPyMhIli5dSnx8PC1btuTrr7+2WUiRMYqiMKyinsqeCj22m6m5ysSqlgZq+EqRJYQQWvnomIVYE3ySDW0Znsa91SvEHdxG9Pqf8Or9ts0fPz+zWkqXK1eOZs2aAeDg4MDs2bO5ceMG4eHh/Pbbb/j5+dkspMichoUezssq4KTQYJ2JxRdkXpYQQmjhQpTK7FMWJlTVUUSDK3DoHF1wb9+XuMM7eHD1rM0fPz+TVuB5VDFXhb0d9fQsqdBnl5kx+82YZF6WEELY1LhgM4WcYVwV7X7dutRuhV3hkkStmo9qkS/ctiIFVh7maFD4qbGe2XV1zDltofVGM6EJUmQJIYQtbLthYe2/Kp8H6XHKxrYMT6Po9Hh0G0riv+eIO7pTsxz5jRRYeZyiKIysrGdbOz0nwlVqrTLxd5gUWUIIkZ1MFpXR+83UL6DwUknt58E6lK6CU9UGRK/7CcsDadtgC1Jg5RNNCj+cl+XpAHXXmPjPJRkmFkKI7PL9OQunImCWDdsyPI2x40DM96OI2fG71lHyhVQF1pgxYwgJCQFgz549KX2wRN5Q3E3hz04GupVQeGWHmXeCzZhlXpYQQmSpiAcqHxy20K+sQk3fnDOOYfAphFuTrsTs+B1TxF2t4+R5qd752bNnc+vWLQCaNm3KmTNnNAklso+zQeGXJnpm1tYx46SFdpvMhMu8LCGEyDLTjlp4YIHptWzfluFp3Fq+jM7Jhai1P2odJc9LVWAVKFCA4OBgAFRVzTHDmiJrKYrCmCp6NrfVczhUpdZqEyfDpcgSQojndT5S5evTFt6rpqOQc877HapzdMbYvj/xx3bz4IoMomSnVAVW7969eeutt9Dr9SiKQp06ddDr9Wn+MRis9igVuUSLIg/nZbnaPZyXtfyyzMsSQojnMfaAmaIuMLpyzjk0+DjnWi2wK1qayJX/J20bslGqKunzzz+nRYsWnDlzhjFjxjBixAj8/f21yiZsoIS7wr5OBgbsMfPidjPvhal8WEOHXpfzvnkJIUROtjnEwoYQld+b63HUsC3D0yg6HR5dh3Lv63HEHd6OS1BLrSPlSU8MQ7Vq1YpWrVqxZs0ahg0bRvny5bXIJWzIxU7hP830BHpbePeQheNhKkua6vFwyLkfEEIIkZOYLCqjD5hpVFChe4mc/9npUKoyTtUaEbX+J5yqNkDn4KR1pDzH6hjmzp07U4orVVWJiYlBVWWeTl6lKArvVNPzRxs9++6oBK02cTZC3m8hhMiIeWctnIuEWXX1uWb+srHTQCxxMcRs+03rKHlSugeJd+/eTbNmzXBycsLDwwMnJyeaN2/O3r17bZVP2FibYjoOdTFgr4faa0ysuSrH54UQIj3hCSqTj1gYUE6huk/uKK4ADF4FcGvanZidKzCF3dY6Tp5jtcDaunUrLVq04M6dO7z77rt8++23TJgwgTt37tC8eXO2bdtmy5zChkobFfZ3MtCyiEKXrWamHDFjkdFLIYRI05SjFpIs8HHNnNeW4WncWvRE5+xG1Dpp25DVrJ4KOHHiRNq1a8fq1atTDXdOnjyZLl26MHHiRFq0aGGTkML23OwVfm+hZ/oxC5OOWDgWpvJLEz3u9rnn25kQQmS3MxEq356xML2WjgI5sC3D0+gcnDB2GEDErzN4cOkkDqVe0DpSnmF1BOvkyZMMGzbsiWPJiqIwbNgwTpw4ke3hhLZ0isLEQD1rW+nZdVOl9hoT/0TKSJYQQiQbe8BMgBu8lYPbMjyNc81m2PmXI3LVPFSLWes4eYbVPcLV1ZUbN26kue769eu4urpmWyiRs3QoruNgFwOqCrVWm9hwTeZlCSHEH9csbLquMqO2Hgd97hu9SvawbcMQkq5fIu6gTP/JKlYLrE6dOjFhwgQ2b96cavmWLVt4//336dy5c7aHEzlHOQ+F4C4GGhdS6LjZzMfHzHJWqRAi30qyqIw5YKZpIYXOxXNvcZXMoURFnAKbELVhIZaEWK3j5AlWC6wvvviCkiVL0rZtWzw8PChXrhweHh60bduWEiVK8MUXX9gyp8gBjPYKq1vpmRSoY+JhCy9uN3M/SYosIUT+881pCxeic1dbhqcxdhyImhBHzFZp25AVrE5y9/T0ZP/+/axfv54///yTiIgIvLy8aNCgAe3bt0eny73Hm8Wz0ykKU2roqeat0GeXmTprTKxuaaC0MW98wAghxNOEJqhMPWphUDkdVbzzzmefwdMX12Y9iNm2DJe6bTD4FNY6Uq6W7gUFdTodnTp1olOnTrbKI3KJLgE6gjsrdNliotZqE/9ppqd1MSm6hRB53+QjFiwqTKuZ9z7z3Jq/SFzwZqLW/oj3gA+0jpOr5b29Q9hMRU+Fg10M1C2g0G6zmc+Oy7wsIUTedipcZd5ZC5MCdfg65Z3Rq2Q6e0eMHQYQf+IvEi78rXWcXE0KLPFcPBwU1rXSM6GqjgmHLLy8w0yszMsSQuRBqvrweoOl3GBEpbz769OpRlPsi5cnatV8advwHPLuHiJsRq9T+LiWnt+b69lwTaXeWhNXoqXIEkLkLeuvqWy7oTKzjh77XNyW4WkURcGj2zCSbl4m9sDmp99ApEkKLJFlepTUsb+zgftJUHO1ie03pF+WECJvSDSrjD1gpmURhQ7+ebe4SmZfvBzONZsT/cfPWOKlbcOzSLPASkhIYMSIERw8eNDWeUQu94KXwqEuBmr4KLTaaObLEzIvSwiR+3192sKlGPiyTt5py/A0xg79URMTiN7yq9ZRcqU0CyxHR0d++ukn4uPjbZ1H5AFejgp/tNEz9gUdY4Mt9NllJt4kRZYQIne6G6/y4VELQyvoqOyVP4orAL2HD27NX+L+njUk3Uv7yi7COquHCOvVq0dwcLAts4g8xKBT+Ly2nqXN9Ky8otJgnYlr96XIEkLkPpMOW9ApMLVG/ptV49a0B3p3L6LWfK91lFzHah+sDz/8kN69e2MwGGjbti1+fn5PDIt6eXlle0CRu71cSkd5j4f9smqsMvF7cz1NCue/DykhRO50Ikzl+/MWvqyjw8cx/4xeJVPsHTB2HEj4ok9IOH8Ux3KBWkfKNdIdwbp8+TLjxo2jcuXK+Pn54evrm+qPEBlRzVvhcFcDVbwUWvxh5utTMi9LCJHzqarKqP1myrjDGxXz7xdDp+qNsC9RkcjV36GapW1DRlkdwVqwYEG+mcgnsp+Po8LmtnreDrYwcr+Fo2Eq/1dfj6NB9jEhRM60+qrKzlsqG1rrsdPl388qRVHw6DqUu1+OJHb/RlwbdNA6Uq5gtcDq16+fDWOI/MCgU/iyrp7qPgqD95o5HWFmZQs9RV3z7weXECJnemBWGRdspk1RhXb++Xf0Kpm9f1mcg1oSvXERzoGN0Tm7aR0px3vqXhMREcHevXv59ddfiYiIAB62cbBYpMeReDZ9yuj4s6OB23EqNVab+PO27EtCiJxl9ikL/95/2JZBPGRs3w81KZHozdK2ISOsFlhms5n33nuPYsWK0bhxY/r06cOVK1cA6NatG9OmTbNZSJH31PB9OC+rvFGh6Xoz887IvCwhRM5wJ07lo2MW3qioo4KnjLAn0xu9cWv5Mvf3riXpTojWcXI8qwXW5MmTmTt3Ll988QVnzpxJ9cuvU6dOrFu3ziYBRd7l56Swrb2eoRV0DPvLwuC9Zh6YpcgSQmjr/cNm7HQwJVAODT7OrUk39B4+0rYhA6zuPQsXLmT69OkMGzaMMmXKpFpXqlQpLl26lO3hRN5np1P4ur6eBY30LLqg0mS9mZuxUmQJIbRxLFRlwXmVqTV0eOXDtgxPo9jZY+z0OglnDpJw9rDWcXI0qwVWWFgYFSpUSHOdxWIhKSkp20KJ/Kd/OR17Ouq5dl+l5moT++/IvCwhhG0lt2Wo4AFDK8jolTVOVRtgX+qF/7ZtMGkdJ8eyugeVLVuWrVu3prlu586dVK5cOdtCifyptp+OI10NlHBTaLzezA/npMgSQtjOiisqe26rfFVXjyEft2V4modtG4ZguhtC7L4/tI6TY1lt0zB69GgGDRqEnZ0dPXr0AOD69evs37+fOXPmsHDhQltlFPlIQWeFne31jNxnYdBeM0dDVWbV1WGvlw87IUT2STCpjA82076YQquiMnr1NPZFS+NcuxXRG3/BObApOhdp2/C4dPtghYeHM2XKFKZPnw5Aly5dcHZ25qOPPuKll16yWUiRv9jrFeY11BPoozB8n5mT4SrLW+gp4CxFlhAie3x1ysL1WNjUVtoyZJSxfT/ij+0letMveHR/Q+s4OY7VAgtgzJgxDB48mL/++ouwsDC8vLyoV68e7u7utson8rHBFXRU9oLuW83UWG1iVUs9tXzlm6UQImvdilP5+JiFEZV0lPOQL3IZpXfzxL3VK0Rt+AmX+u2xK1hc60g5ylN/W7m6utK6dWt69epFmzZtpLgSNlWvwMN5WUVdFBquM/PzPzIvSwiRtd47ZMZRD5OkLUOmuTbujN6zAJGrpW3D49Ldm0JDQ5k4cSItWrSgUqVKtGjRgokTJ3Lv3j1b5ROCwi4KuzvoebWUQr/dZt7aZybJIq0chBDP7/A9Cwv/UZlWU4eHg4xeZZZisMej8+s8OHeY+DMHtY6To1gtsIKDgylTpgxz5szB1dWV+vXr4+rqypw5cyhdujTBwcG2zCnyOQe9wg+N9HxTX8e3Zyy0+sPMvXgpsoQQz+5hWwYLlT1hUHkZvXpWji/Uw6FMVaKkbUMqVveoN998k0qVKhESEsLq1av57rvvWL16NSEhIVSqVInhw4fbMqcQKIrCGxX1bG+v50zkw35ZR0OlyBJCPJtll1X+uqMyS9oyPBdFUTB2HYLp3k3u75WrvCSzWmCdPn2aCRMmYDQaUy03Go1MmDCBU6dOZXs4IdLSqJCOw10M+Doq1F9rYslFmZclhMiceJPK2wfNdC6u0LyIjF49L/vCJXGp24bozUsw34/SOk6OYHWvKl26NJGRkWmui4qKomTJktmVSYinKuaqsLejnpdKKvTeaWbcATMmmZclhMigGScs3IqDL2pLW4as4t72NVAtRG/6ResoOYLVAuuLL75g8uTJ7N69O9XyXbt2MWXKFGbMmJGlQQICAihfvjzVqlWjWrVq/PbbbynrLly4QL169ShbtixBQUGcOXMmSx9b5E5OBoWFjfXMqqtj1ikLbTeZCUuQIksIkb4bsSqf/m3hrco6yhjl0GBW0bt54N76VWL/+oOkW1e1jqO5VH2wXnjhBRTlfztbVFQUzZo1w2g04uvry71794iKisLT05N33nmHtm3bZmmY5cuXp3kJniFDhjB48GD69evH8uXLGThwIPv378/Sxxa5k6IovFVZTxUvhRe3mam12sSqlgaqesuHphAibe8eNONigInV5dBgVnNt2InYfX8QuWo+PsOmp6op8ptUBVaNGjVSvRg1atSweaDH3b17l6NHj7JlyxYAunfvzvDhw7l69SoBAQHahhM5RtPCOg53Vei61US9tSZ+aqTnpVLy4SmESC34roVfLqp811CP0T7//vLPLorBDmPnQYT9MIWE0wdwqlxX60iaSVVgaX19wVdffRWLxULt2rX55JNP8PX1JSQkhMKFC2MwPIyqKAr+/v5cu3bNaoE1fPhwjEYj3bp1o3v37hl+/IiIiKx4GkIj7sC6BjDqqCM9d8Bf1+OYWDmRvHgZQ9lXRW6SU/ZXVYXhe52pbIQuvjGEh2udKG9SC5VBV6Iy4Svn4VigFIo+3YvG5CjPs696eXml+jnHPOs9e/bg7+9PUlISEydOpG/fvvzxx8OrdD8+xKiq6c+zmTt3LoGBgc+U4/EXSOQuXsDvrVW+PGnh7YMOnI9zZGkzPZ55sIGg7KsiN8kJ++uSixYOh5vZ2V6Pr7ej1nHytKSXhnPn8zewP7UXt6YZH+jICbJqX023wDpw4AC///47ISEhJCQkpFqnKApr1qzJkhAA/v7+ANjZ2TFq1CjKli0LQLFixbh+/TomkwmDwYCqqoSEhKRsL8TjFEVhbJWH87J67ng4L2tNSwOVvPJekSWEyJjYJJV3DprpFqDQpLBMH8hudoUCcKnfjujNS3Cu2Ry9m4fWkWzO6l42e/Zs6tWrx5IlS7hx4wYxMTGp/kRHR2dZiNjY2FQtIZYuXUr16tUB8PPzo3r16ixevBiAFStWEBAQIPOvxFO1LPqwX5azAWqvMbHyivTLEiK/+uKEhXvx0pbBltzb9AFFR/QfP2sdRRNWR7BmzJjB8OHDmTVrFjpd9lb7d+7coXv37pjNZlRVpWTJkixatChl/fz58+nXrx/Tp0/H3d2dn3/On2+WyLyS7gr7Oxnov9tM921mJlZXmVpDhy4fn9kiRH4Tcl/l878tjHlBR0l3+b9vK3pXI+6tXyVqzfe4NOiIfZH81T/TaoEVFxdH586ds724AihZsiTHjh2zur5cuXLSlkE8Mxc7hd+a6wn828J7hywcD1NZ3FTOIBIiv3jnoBl3e3ivmhwatDXXhh2J3beBqNXz8Xnj03zVtsHq3vbSSy+xceNGW2YRItsoisKEano2tNGz97ZK0GoT5yKlKakQed2+OxaWXlKZXkuPm3ypsjlFb8DYZTAPLvxNwsl9WsexKasjWLNmzWLgwIH06tWLFi1a4OHh8cQ23bp1y85sQmS5tsV0HOqi0GWLiaDVJhY31dOpuHyrFSIvsqgqo/ZbCPSBfmWluNKKU8UgHMrXJHLNDzhWrIVisNc6kk1YLbDOnDnDn3/+ybVr1/jPf/7zxHpFUTCbzdkaTojsUMaocKCzgdd2mem8xczUGioTq8u8LCHymsUXVA7dU9nTQS//vzXm0WUQdz4fxv3da3Br/qLWcWzCaoE1cOBA3N3dWbduHWXLlsXePn9UnCJ/cLNXWNFSz8fHLEw6YuFYqMrPTfS4yyEEIfKE+0kqEw6ZebGEQsNCMkqtNbuCxXGt34HoLUtxDmqB3s1T60jZzmqBde7cOVauXEmbNm1smUcIm9EpCh8E6qnmrfDqTjN11phY3dJAWQ8psoTI7T7720L4A/hc2jLkGO5t+hB3ZCdRGxbi9fJoreNkO6tlfZUqVbh7964tswihiY7FdRzsYsCiQtAaE39ck35ZQuRm/8aozDhhYdwLOgLc5AtTTqFzccO9bR/igreQeP2i1nGyndUC69tvv+Wrr75i69atmEwmW2YSwubKeygEdzHQsKBCh81mph8zP/WSTEKInOntg2Y8HWCCtGXIcVzqtcPgV4zIVfPz/Ges1b2vcePGnD17ljZt2uDk5IS7u3uqP0aj0ZY5hch2RnuFNa30fBCo4/3DFl7abuZ+Ut7+ABAir9l7y8Kyyyqf1NLjaiejVzmNojfg0WUwiZdOEv/3n1rHyVZW52CNHTs2XzUEEwIezsuaWkNPNS+F13abqbvGxOpWBkpJ92chcjyLqjLqgJlavgp9ysj/2ZzKsUJNHCsGEbX2B5wq1Uaxy5sn0VktsKZMmWLDGELkLF1L6DhgVOiy1USt1Sb+00xPq6JyuEGInOznf1SOhsJfnaTtSk5n7DyIO58NJWbXStxbvqx1nGwhvzGEsKKSl8LBLgbq+Cm03WTmi79lXpYQOVVMosp7h8y8UkqhXgH51ZbT2RUohmvDTsRs/Q/mqDCt42QLqyNYzZo1e+qNd+zYkaVhhMhpPB0U1rXS88ERC28ftHA0TOXHRnqcDfLtWIicZPpxC1GJ8GmQtGXILdxb9yLu8PaHbRt6jdU6TpazWuY7Ozvj4uKS6s+DBw84dOgQFy5cwNXV1ZY5hdCMXqcwvZaeZc31rP1Xpf5aE1djZCRLiJziSrTKlyctjK+qw99VvvzkFjpnN9zbvkbcwa0kXvtH6zhZzuoI1vr169NcfvfuXTp16kSfPn2yLZQQOdGLJXWU+++8rJqrTCxrrqdZETkUIYTWxh804+sEb1eR/4+5jUvdttz/az2Rq+bhO3Jmnjq5LtN7o5+fH++88w4TJ07MjjxC5GhVvBUOdTEQ6KPQaqOZWSdlXpYQWtp9y8KKKyqfBelxkbYMuY6i1z9s23DlDPHHdmsdJ0s9U7lvsVi4detWVmcRIlfwdlT4o42e0S/oGH3AQt9dZuJNUmQJYWtmi8pb+8zU9lN4pZQUV7mVY7lAHCvXIWrtj6iJD7SOk2WsHiI8evToE8sSExM5e/YsU6dOJSgoKFuDCZGTGXQKX9TWU91b4fU9Zs5EmlnZUi/zP4SwoQX/qPwdDgc6S1uG3M7YeRB3Ph1CzM7luLd+Ves4WcJqgVWzZs0njoUmHwqpXbs233//ffYmEyIX6FVaRwUPha7/nZe1vIWeRoVkHogQ2S0qUeX9Q2Z6l1ao7Sf/53I7O98iuDbqTMz2ZbjUbo3ew0frSM/NaoG1c+fOJ5Y5OjpStGhRihQpkq2hhMhNqvsoHO5q4KVtZppvMPNVXZU3K+ry1GRNIXKaj49ZiDVJW4a8xL1VL+IObSNq/QK8er+tdZznZrXAaty4sS1zCJGr+TgqbGmnZ3ywhRH7LBwLVfmmvh5H6ZclRJa7GKUy65SFD6rrKOIi/8fyCp2TC+7t+hK5bA4uDTrhEFBe60jPRcZVhcgiBp3CV3X1/NxYz5JLKo3Xm7kRK5Pfhchq44LNFHSCsdKWIc9xqdMau8IliVo1L9efoZ1q73Rzc8Pd3T1Df4xGo1aZhcjRXiur48+Oem7GqdRYZeKv2xatIwmRZ2y/YWHNvyqf15YrKuRFik6PsesQEv89R/yRJ6cq5SapDhGOHTv2qfNG/vzzT7Zv3y7zS4RIR01fHYe7KLy43UzTDWa+rqcypILMFRHieZgsKqMPmKlfQKFnSfkdlFc5lqmKU5X6RK1bgOML9dA5OGod6ZmkKrCmTJlidcO//vqLKVOmsH37dqpUqcKkSZOyO5sQuVoBZ4Vt7fSMOWBh6J8WjobCnHo6HPTyi0GIZ/HDOQsnw+FQFzmJJK8zdnqd258MJmbH7xjb5s4rxzz1APbevXtp0aIFjRo1IiwsjJUrV3L8+HG6detmi3xC5Gr2eoW59fX80FDPwn8sNNtg5lZc7p5XIIQWIh+ofHDEQt8yCjV9Ze5VXmfwKYRbk67c37EcU8Q9reM8E6t76Z49e2jWrBmNGzcmMjKSVatWcfToUbp06WLDeELkDQPL69jdQc/VGJWaq0wcuCPzsoTIjGnHLMSbYHotOdSeX7i1fBnF0ZmodT9qHeWZPFFg7dq1i6ZNm9KkSRNiY2NZt24dhw8fplOnTlrkEyLPqFNAx+GuBoq7KjReb2bBeSmyhMiIfyJV5pyy8F41HYWlLUO+oXN0xti+H/FHd/Hgyhmt42RaqgKrSZMmNG/enAcPHvDHH38QHBxM+/bttcomRJ5TyFlhZwc9/crqGLjHzPC/zCRZ5JChEOkZG2ymiAuMeUEODeY3zkEtsCtaishV81AtuetLaapJ7nv27AHg5MmTvPTSS+neUFEUoqKisi+ZEHmUg15hfkM9gT4wYp+FE+Eqy1vo8XOSb+ZCPG7LdQvrr6ksay6Ne/MjRafHo+sw7n09jrjDO3AJaqF1pAxLVWBNnjxZqxxC5DtDKuip7KnQfZuZGqtMrGqpl8m7QjzCZFEZvd9Mw4IKPUpIcZVfOZSqjFO1hkStX4BT1froHJy0jpQhUmAJoaH6BXUc6arQbauZBuvMfNfgYaNSIQTMP2vhbCQc6aqXtgz5nLHjQG5/MoiYbb9hbN9P6zgZIp/kQmisiIvC7g56epVS6LvbzKj9Mi9LiPAElUlHLAwop1DdR4qr/M7gXRC3pt2J2bkCU/gdreNkiBRYQuQAjgaFHxvpmVtPxzenLbT+w0xoghRZIv+aetRCkgU+qiltGcRDbi16onN2I2pt7mjbIAWWEDmEoii8WUnPtvZ6TkU87Jd1LFSKLJH/nI1Q+eaMhfer6SjoLKNX4iGdgxPGDv2JP76HB5dOaR3nqaTAEiKHaVzoYb8sH0eF+mtNLL2Yu05NFuJ5jQ02U9wVRklbBvEY55rNsStWJle0bZC9V4gcyN9VYW9HPd1LKPTaaWZ8sBmTzMsS+cDGEAsbQ1Rm1NbLdTvFExSdDo+uQ0m6fpG4g1u1jpMuKbCEyKGcDAqLmuj5qo6Or05aaLfJTLjMyxJ5WJJFZcwBM00KKXQJkOJKpM2hZCWcqjcmasNCLAlxWsexSgosIXIwRVEY9YKezW31HA1VqbnaxIkwKbJE3vR/Zyz8EwWz6kpbBpE+Y6eBqAlxxGz9j9ZRrJICS4hcoHmRh/Oy3O2g7loTS64auByt8sAsxZbIG8ISVKYctfB6OR1VvaW4EukzePrh2qwHMbtWYQq9pXWcNBmevokQIicIcFPY19nAwD1mRh5xYuQREwC+jlDUBYq6KP/98/DfxVwf/l3EBZzlEiMih5t8xILZAtNqyvd+kTFuzV4k7sBmotb+gPeAD7SO8wQpsITIRZwNCr821TO6VCTRBiPXY+H6ffXh37Eq++5YuB4LYQ9S387LIe0i7NG/3eylCBPaOB2uMu+shc+CdHJNTpFhOgdH3Dv2J2LxFyRc+BvHMlW1jpSKFFhC5DKKolDaTcXLy/o3/XiTyo1YCIn9X/GV/PfhUAur/4W78alv425H6sLLVaHYY0WY0R6ZGyOylKqqjD5gpoQbjKgko1cic5wDmxK7dx1Rq+bjMO5rFF3OaUwrBZYQeZCTQaG0EUobrRdDD8wqNx8rvpL/PhUBm65buBUHj87ycjGQ5uhXUReFoq4P/+3tIEWYyLgN11S23lBZ00qPvbRlEJmk6HQYuw7h3qzRxAZvwbVuW60jpZACS4h8ykGvUMIdSrhb/6WWZFG5HfffkbD7qYuwf6Jgx00LN+Pg0bn2DvrURVjqUbCH//Z1Ap0UYfleovlhW4YWRRQ6+sv+IJ6NQ0AFnGs0JXrDQpyrNULn5KJ1JEAKLCFEOux0DyfLF3NVoEDa25gtKnfi0x4Ju3aflHlhSZZH7xeKOCePfKU9IlbQCfQ6+aWbl809beFSDKxsKW0ZxPNx7ziA+JP7iN66FI9Or2sdB5ACSwjxnPQ6hcIuUNhFIcjKNhZVJTQBrsdCyH31iWLs8L2HRViC+ZH7VaDQf4uwYk8cjnz4dyHnh0WgyH3uxat8eMzCkPI6KnvJeyiej8HDF7dmLxG9dSmuddth8C2sdSQpsIQQ2U+nKPg5gZ8TBPqk/ctUVVXCH/BI4ZW6CDsZYSHkPsSa/ncbBSjobGVOmMv/2lTIJVdynklHLCjAh9KWQWQR12bdiT2wicg13+Pz+mSt40iBJYTIGRRFwdsRvB2x2mhSVVWiEnniUGTyv3fcfDgSFpWY+nbJvcKKuabdpkJ6hdnWiTCV785ZmFlbh4+jvO4ia+jsHTF2Gkj4ok9J+OcYjmWra5pHCiwhRK6hKAoeDuDhQLqHlWISVW7EPVKEPdIr7K/b0itMS8ltGcq4w5vSlkFkMafqjbHfu5bIVfMpMO4bFL12bRukwBJC5Dlu9grl7aG8h/WCKO6/vcLSmpyfXq+wYk9MypdeYZmx5l+VHTdV1rfWy/w5keUURcGj6xDufvkWsQc24Vq/vWZZpMASQuRLzgaFMkYo84y9wk6Gw8YQ6RWWGQ/MKuOCzbQuqtCuWP57/sI27P3L4VyrBdF/LMK5emN0zq6a5JACSwghrMhor7BbyYcjM9grzFGfThGWh3uFzTll4WoMrG0lbRlE9jJ26E/8338SvXkJHl2HaJJBCiwhhHgOdjoFf1fwf8ZeYf/eh7+eoVdYMReFArmoV9idOJVpxyy8UVFHRc/ckVnkXnqjN24tehK9aTEu9dtj51fU5hmkwBJCiGyW0V5h9+Ktt6mw1iussLU2FTmsV9jEw2YMOpgcKBPbhW24NelG7P6NRK35Hp9BU23++FJgCSFEDqBTFAo4QwFnqOH7bL3CToQ/LMIy2issuYFr4WzuFXYyUseP51Vm19XhLW0ZhI0o9g4YO79O+MLpJJw7gmP5GjZ9fCmwhBAil8jOXmF+Tum3qXjWXmGqqvL+3w6U94ChFWX0StiWU9WG2JesTOTq+RQY/382bdsgBZYQQuQhWdEr7M//9goLz2CvsEdbV7japX7MlVdV/go1sKmNtGUQtve/tg0jid23AdeGnWz22FJgCSFEPpRdvcKM9qnPkNxyXaVlQROti9ll8zMSIm32xcrgHNSS6I2/4BzYFJ2Lm00eVwosIYQQacqKXmEudvBRlQTAyXbBhXiMsX0/4o//SfTmxXh0G2aTx5QCSwghxDPLSK+w8HDV6johbEHv7oVby5eJ/mMhLvXaY1fQP9sfU2YcCiGEECLPc2vcBb2nH5Grv7PJ40mBJYQQQog8T7Gzx6PT6zw4d5j4M4ey/fGkwBJCCCFEvuBYpT4OpasQtfo7VLPp6Td4DlJgCSGEECJfUBQFY9ehmO5d5/6f67P1saTAEkIIIUS+YV+kJC512hC9aTHm+1HZ9jhSYAkhhBAiX3Fv1xdUC9GbFmfbY0iBJYQQQoh8Re/mgXurXsTu20DSravZ8hhSYAkhhBAi33Ft1BmDV0EiV3+HqmZ9rzYpsIQQQgiR7ygGO4ydB/Hg/FESTgdn+f3nigLrwoUL1KtXj7JlyxIUFMSZM2e0jiSEEEKIXM6xch0cylYjas33qKakLL3vXFFgDRkyhMGDB/PPP//w9ttvM3DgQK0jCSGEECKXUxQFjy5DMIXe4v7etVl63zm+wLp79y5Hjx6ld+/eAHTv3p0rV65w9epVbYMJIYQQItezK1wCl3rtiN78K2pcdJbdb46/2HNISAiFCxfGYHgYVVEU/P39uXbtGgEBAWneZvjw4RiNRrp160b37t0z/FgRERFZEVmIbCf7qshNZH8VOZ1atwN2fv5EJiShhIc/0314eXml+jnHF1jwsKh61NNm+8+dO5fAwMBneqzHXyAhcirZV0VuIvuryNG8vKBoAIbw8CzbV3P8IcJixYpx/fp1TKaH1wxSVZWQkBD8/f01TiaEEEIIkbYcX2D5+flRvXp1Fi9+2G11xYoVBAQEWD08KIQQQgihtVxxiHD+/Pn069eP6dOn4+7uzs8//6x1JCGEEEIIq3JFgVWuXDn279+vdQwhhBBCiAzJ8YcIhRBCCCFyGymwHrFixQqtIwiRIbKvitxE9leRW2TlvioF1iNWrlypdQQhMkT2VZGbyP4qcous3FdzxRysjIqPjwfg7Nmzz3T7qKgojh49mpWRhMgWsq+K3ET2V5FbPO++Wr58eZydnQFQ1Kd17cxFlixZknJJHSGEEEIIWzpy5EhKo/M8VWCFhoayefNmAgICcHJy0jqOEEIIIfKRPDuCJYQQQgiRE8gkdyGEEEKILCYFlhBCCCFEFpMCSwghhBAii0mBJYQQQgiRxaTAAkaOHElAQACKonDq1Cmt4wiRpoSEBLp06ULZsmWpVq0abdq04erVq1rHEiJNrVq1okqVKlSrVo2GDRty/PhxrSMJka6pU6dmaR0gBRbQo0cP/vzzT4oXL651FCHSNXjwYM6fP8/x48fp0KEDgwcP1jqSEGlatmwZJ06c4Pjx44wdO5YBAwZoHUkIq44ePcqBAwfw9/fPsvuUAgto1KgRRYsW1TqGEOlydHSkXbt2KIoCQJ06dbh8+bLGqYRIm4eHR8q/o6Ki0Onk143ImR48eMCbb77Jt99+m/L5mhXy1KVyhMhP5syZQ8eOHbWOIYRVr732Gjt37gRg06ZNGqcRIm2TJk2id+/elChRIkvvV75SCJELTZ8+nQsXLvDxxx9rHUUIqxYtWkRISAgfffQR48eP1zqOEE/Yv38/hw4d4o033sjy+5YCS4hcZsaMGaxcuZKNGzemXJJBiJysb9++7Ny5k7CwMK2jCJHK7t27OXfuHCVKlCAgIIDr16/TunVrNm7c+Nz3LQWWELnIl19+ydKlS9m6dWuqOS5C5CTR0dHcvHkz5edVq1bh7e2Nl5eXhqmEeNKECRO4efMmV69e5erVqxQtWpTNmzfTtm3b575vmYMFvPnmm6xZs4bbt2/TokULXF1duXjxotaxhEjl+vXrjB07lpIlS9K0aVMAHBwcCA4O1jiZEKlFRUXRvXt34uPj0el0+Pr6sn79+iydQCxETicXexZCCCGEyGJyiFAIIYQQIotJgSWEEEIIkcWkwBJCCCGEyGJSYAkhhBBCZDEpsIQQQgghspgUWEIIIYQQWUwKLCGETU2ZMgVFUVL++Pr60rx5c/bu3atZJkVRmDFjRqZuc/z4caZMmUJcXFyq5bt27UJRFA4fPpyVEYUQuYwUWEIIm3NycmL//v3s37+f//u//yMsLIzmzZtz8uRJraNl2PHjx5k6deoTBVZgYCD79++nQoUKGiUTQuQE0sldCGFzOp2OOnXqpPwcFBREQEAA8+fPZ+7cuRome37u7u6pnpsQIn+SESwhhOb8/f3x8fHhypUrWCwWpk+fTokSJXBwcKBMmTLMmjUr1fZTpkzB1dWVQ4cOERQUhKOjIxUqVGD9+vWptgsICGD48OGpli1fvhxFUbh69arVPBs2bKBly5b4+fnh7u5O7dq12bRpU8r6hQsX0r9/fwB8fX1RFIWAgAAg7UOECQkJjB07liJFiuDg4MALL7zAr7/+muox+/XrR+XKldm1axfVq1fHxcWFoKAgjhw5ktGXUQiRg0iBJYTQXHR0NOHh4RQuXJjx48fzwQcf0Lt3b9atW0eXLl0YPXo006ZNS3WbpKQkevbsSd++fVm5ciWlS5ema9eunDp16rnzXLlyhY4dO/LLL7+wYsUK6tevT7t27di1axcA7du3Z+LEiQBs2rSJ/fv3s2rVKqv39+qrr/Ltt98yZswY1q5dS82aNXn11Vf55ZdfUm13+/ZtRo4cyfjx4/ntt9+Ii4uja9euJCUlPfdzEkLYlhwiFEJowmQyAf+7iLXZbKZFixb06dOHsWPHphRUrVq1Ijo6ms8++4zRo0fj6uoKQGJiIhMnTmTAgAEAtG7dmtKlSzN9+vQnRocy69FRL4vFQtOmTTl9+jTfffcdTZo0wdfXl1KlSgFQo0YNfHx8rN7XiRMnWLlyJd988w1vvPFGStabN2/ywQcf0KdPn5Rtw8PD2b17N5UqVQLA0dGRli1bEhwcTIMGDZ7rOQkhbEtGsIQQNhcbG4udnR12dnaUKFGCnTt3MnfuXFxdXVNGph71yiuvEBsby7Fjx1It79q1a8q/9Xo9nTp14sCBA8+d7/r16/Tt25ciRYpgMBiws7Njy5Yt/PPPP5m+r+SzI9N6Tv/++y8hISEpywoXLpxSXAFUrFgxJY8QIneRESwhhM05OTmxZ88eFEXBx8eHYsWKodPpWLx4MQAFCxZMtX3yz+Hh4SnL7Ozs8PT0TLWdn58ft27deq5sFouFTp06ERUVxYcffkjp0qVxcXFh0qRJXLt2LdP3FxERgcFgwNvbO9XyR59TsWLFAPDw8Ei1jb29PfBwDpcQIneRAksIYXM6nY6aNWs+sdzLywuAO3fuUKRIkZTlt2/fTrUeHs7BioiISFVk3b17l0KFCqX87OjoSGJiYqrHeLRIS8vFixc5duwYq1evpnPnzinL4+PjM/LUnuDl5YXJZCI8PDxV/rSekxAi75BDhEKIHCMoKAg7OzuWLVuWavlvv/2Gi4sLgYGBqZY/OrHcbDazdu1aateunbKsaNGinD17NtVttm7dmm6G5EIqefQI4N9//+Wvv/5KtV1GR5eS506l9ZyKFy+eMnolhMhbZARLCJFj+Pj4MHLkSGbMmIGDgwP169dn+/btzJ8/n6lTp+Li4pKyrb29PR999BEJCQmUKFGCb7/9luvXr/Puu++mbNOjRw+GDRvG1KlTqVevHhs2bODgwYPpZihfvjxFixZlwoQJmM1mYmNjmTx5cqoRNSClkeg333xDly5dcHZ25oUXXnji/qpUqUL37t0ZM2YMcXFxVKpUiWXLlrFp0yYWLVr0PC+XECIHkxEsIUSO8vnnnzN16lR+/vlnOnTowIoVK5g5cyYffPBBqu3s7OxYunQpCxYsoEuXLly4cIEVK1ZQpUqVlG1ef/11xo0bx//93//x4osvkpCQwEcffZTu4zs4OLBy5UocHBx48cUX+eCDD3jvvfdo3Lhxqu2qV6/OlClTWLx4MfXq1aNjx45W73Px4sUMGzaMGTNm0LFjR4KDg1m8eHGqMwiFEHmLoqqqqnUIIYTIjClTpjBjxgzu37+vdRQhhEiTjGAJIYQQQmQxKbCEEEIIIbKYHCIUQgghhMhiMoIlhBBCCJHFpMASQgghhMhiUmAJIYQQQmSx/wcMO0JDQuPq9AAAAABJRU5ErkJggg=="
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "plot(ms_emu_result, title=\"Model selection result - emulation\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can see that emualtion also correctly identified the first model is as the preferred one."
   ]
  }
 ],
 "metadata": {
  "@webio": {
   "lastCommId": null,
   "lastKernelId": null
  },
  "kernelspec": {
   "display_name": "Julia 1.7.2",
   "language": "julia",
   "name": "julia-1.7"
  },
  "language_info": {
   "file_extension": ".jl",
   "mimetype": "application/julia",
   "name": "julia",
   "version": "1.7.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}