Semi-Gradient-TD(0)-with_State_Aggregation_Assignment.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "markdown",
     "checksum": "cf073b8339c6caf337498b984d0678f8",
     "grade": false,
     "grade_id": "cell-4492106616b7ddd6",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "source": [
    "# Assignment 1 - TD with State Aggregation\n",
    "\n",
    "Welcome to your Course 3 Programming Assignment 1. In this assignment, you will implement **semi-gradient TD(0) with State Aggregation** in an environment with a large state space. This assignment will focus on the **policy evaluation task** (prediction problem) where the goal is to accurately estimate state values under a given (fixed) policy.\n",
    "\n",
    "\n",
    "**In this assignment, you will:**\n",
    "1. Implement semi-gradient TD(0) with function approximation (state aggregation).\n",
    "2. Understand how to use supervised learning approaches to approximate value functions.\n",
    "3. Compare the impact of different resolutions of state aggregation, and see first hand how function approximation can speed up learning through generalization.\n",
    "\n",
    "**Note: You can create new cells for debugging purposes but please do not duplicate any Read-only cells. This may break the grader.**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "markdown",
     "checksum": "26ba6e0644732c204d2547aad08e2db7",
     "grade": false,
     "grade_id": "cell-3eae2f3977544921",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "source": [
    "## 500-State RandomWalk Environment\n",
    "\n",
    "In this assignment, we will implement and use a smaller 500 state  version of the problem we covered in lecture  (see \"State Aggregation with Monte Carlo”, and Example 9.1 in the [textbook](http://www.incompleteideas.net/book/RLbook2018.pdf)). The diagram below illustrates the problem.\n",
    "\n",
    "![](data/randomwalk_diagram.png)\n",
    "\n",
    "There are 500 states numbered from 1 to 500, left to right, and all episodes begin with the agent located at the center, in state 250. For simplicity, we will consider state 0 and state 501 as the left and right terminal states respectively. \n",
    "\n",
    "The episode terminates when the agent reaches the terminal state (state 0) on the left, or the terminal state (state 501) on the right. Termination on the left (state 0) gives the agent a reward of -1, and termination on the right (state 501) gives the agent a reward of +1.\n",
    "\n",
    "The agent can take one of two actions: go left or go right. If the agent chooses the left action, then it transitions uniform randomly into one of the 100 neighboring states to its left. If the agent chooses the right action, then it transitions randomly into one of the 100 neighboring states to its right. \n",
    "\n",
    "States near the edge may have fewer than 100 neighboring states on that side. In this case, all transitions that would have taken the agent past the edge result in termination. If the agent takes the left action from state 50, then it has a 0.5 chance of terminating on the left. If it takes the right action from state 499, then it has a 0.99 chance of terminating on the right.\n",
    "\n",
    "\n",
    "### Your Goal\n",
    "\n",
    "For this assignment, we will consider the problem of **policy evaluation**: estimating state-value function for a fixed policy.You will evaluate a uniform random policy in the 500-State Random Walk environment. This policy takes the right action with 0.5 probability and the left with 0.5 probability, regardless of which state it is in. \n",
    "\n",
    "This environment has a relatively large number of states. Generalization can significantly speed learning as we will show in this assignment. Often in realistic environments, states are high-dimensional and continuous. For these problems, function approximation is not just useful, it is also necessary."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "markdown",
     "checksum": "0d2b36d11dbd26ec6c6109267c14daa6",
     "grade": false,
     "grade_id": "cell-d0f5297c93921d37",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "source": [
    "## Packages\n",
    "\n",
    "You will use the following packages in this assignment.\n",
    "\n",
    "- [numpy](www.numpy.org) : Fundamental package for scientific computing with Python.\n",
    "- [matplotlib](http://matplotlib.org) : Library for plotting graphs in Python.\n",
    "- [RL-Glue](http://www.jmlr.org/papers/v10/tanner09a.html) : Library for reinforcement learning experiments.\n",
    "- [jdc](https://alexhagen.github.io/jdc/) : Jupyter magic that allows defining classes over multiple jupyter notebook cells.\n",
    "- [tqdm](https://tqdm.github.io/) : A package to display progress bar when running experiments\n",
    "- plot_script : custom script to plot results\n",
    "\n",
    "**Please do not import other libraries** - this will break the autograder.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "code",
     "checksum": "33ccd5de59ab1974ad1b867fd105c664",
     "grade": false,
     "grade_id": "cell-a4c4b2e807873209",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "\n",
    "import jdc\n",
    "from tqdm import tqdm\n",
    "\n",
    "from rl_glue import RLGlue\n",
    "from environment import BaseEnvironment\n",
    "from agent import BaseAgent\n",
    "import plot_script"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "markdown",
     "checksum": "b86a50a5ddd969463cfca105d9f35bdf",
     "grade": false,
     "grade_id": "cell-fa84e2185a338aa6",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "source": [
    "## Section 1: Create the 500-State RandomWalk Environment\n",
    "\n",
    "In this section we have provided you with the implementation of the 500-State RandomWalk Environment. It is useful to know how the environment is implemented. We will also use this environment in the next programming assignment. \n",
    "\n",
    "Once the agent chooses which direction to move, the environment determines how far the agent is moved in that direction. Assume the agent passes either 0 (indicating left) or 1 (indicating right) to the environment.\n",
    "\n",
    "Methods needed to implement the environment are: `env_init`, `env_start`, and `env_step`.\n",
    "\n",
    "- `env_init`: This method sets up the environment at the very beginning of the experiment. Relevant parameters are passed through `env_info` dictionary.\n",
    "- `env_start`: This is the first method called when the experiment starts, returning the start state.\n",
    "- `env_step`: This method takes in action and returns reward, next_state, and is_terminal."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "code",
     "checksum": "a39f70fc723d242c51a64de70ddba135",
     "grade": false,
     "grade_id": "cell-931189bd60e2d453",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "outputs": [],
   "source": [
    "# ---------------\n",
    "# Discussion Cell\n",
    "# ---------------\n",
    "\n",
    "class RandomWalkEnvironment(BaseEnvironment):\n",
    "    def env_init(self, env_info={}):\n",
    "        \"\"\"\n",
    "        Setup for the environment called when the experiment first starts.\n",
    "        \n",
    "        Set parameters needed to setup the 500-state random walk environment.\n",
    "        \n",
    "        Assume env_info dict contains:\n",
    "        {\n",
    "            num_states: 500 [int],\n",
    "            start_state: 250 [int],\n",
    "            left_terminal_state: 0 [int],\n",
    "            right_terminal_state: 501 [int],\n",
    "            seed: int\n",
    "        }\n",
    "        \"\"\"\n",
    "        \n",
    "        # set random seed for each run\n",
    "        self.rand_generator = np.random.RandomState(env_info.get(\"seed\")) \n",
    "        \n",
    "        # set each class attribute\n",
    "        self.num_states = env_info[\"num_states\"] \n",
    "        self.start_state = env_info[\"start_state\"] \n",
    "        self.left_terminal_state = env_info[\"left_terminal_state\"] \n",
    "        self.right_terminal_state = env_info[\"right_terminal_state\"]\n",
    "\n",
    "    def env_start(self):\n",
    "        \"\"\"\n",
    "        The first method called when the experiment starts, called before the\n",
    "        agent starts.\n",
    "\n",
    "        Returns:\n",
    "            The first state from the environment.\n",
    "        \"\"\"\n",
    "\n",
    "        # set self.reward_state_term tuple\n",
    "        reward = 0.0\n",
    "        state = self.start_state\n",
    "        is_terminal = False\n",
    "                \n",
    "        self.reward_state_term = (reward, state, is_terminal)\n",
    "        \n",
    "        # return first state from the environment\n",
    "        return self.reward_state_term[1]\n",
    "        \n",
    "    def env_step(self, action):\n",
    "        \"\"\"A step taken by the environment.\n",
    "\n",
    "        Args:\n",
    "            action: The action taken by the agent\n",
    "\n",
    "        Returns:\n",
    "            (float, state, Boolean): a tuple of the reward, state,\n",
    "                and boolean indicating if it's terminal.\n",
    "        \"\"\"\n",
    "        \n",
    "        last_state = self.reward_state_term[1]\n",
    "        \n",
    "        # set reward, current_state, and is_terminal\n",
    "        #\n",
    "        # action: specifies direction of movement - 0 (indicating left) or 1 (indicating right)  [int]\n",
    "        # current state: next state after taking action from the last state [int]\n",
    "        # reward: -1 if terminated left, 1 if terminated right, 0 otherwise [float]\n",
    "        # is_terminal: indicates whether the episode terminated [boolean]\n",
    "        #\n",
    "        # Given action (direction of movement), determine how much to move in that direction from last_state\n",
    "        # All transitions beyond the terminal state are absorbed into the terminal state.\n",
    "        \n",
    "        if action == 0: # left\n",
    "            current_state = max(self.left_terminal_state, last_state + self.rand_generator.choice(range(-100,0)))\n",
    "        elif action == 1: # right\n",
    "            current_state = min(self.right_terminal_state, last_state + self.rand_generator.choice(range(1,101)))\n",
    "        else: \n",
    "            raise ValueError(\"Wrong action value\")\n",
    "        \n",
    "        # terminate left\n",
    "        if current_state == self.left_terminal_state: \n",
    "            reward = -1.0\n",
    "            is_terminal = True\n",
    "        \n",
    "        # terminate right\n",
    "        elif current_state == self.right_terminal_state:\n",
    "            reward = 1.0\n",
    "            is_terminal = True\n",
    "        \n",
    "        else:\n",
    "            reward = 0.0\n",
    "            is_terminal = False\n",
    "        \n",
    "        self.reward_state_term = (reward, current_state, is_terminal)\n",
    "        \n",
    "        return self.reward_state_term\n",
    "        "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "markdown",
     "checksum": "57c21c07cae1b71e89701a6b2d3fc2b2",
     "grade": false,
     "grade_id": "cell-2e3638a24f9a10be",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "source": [
    "## Section 2: Create Semi-gradient TD(0) Agent with State Aggregation\n",
    "\n",
    "Now let's create the Agent that interacts with the Environment.\n",
    "\n",
    "You will create an Agent that learns with semi-gradient TD(0) with state aggregation.\n",
    "For state aggregation, if the resolution (num_groups) is 10, then 500 states are partitioned into 10 groups of 50 states each (i.e., states 1-50 are one group, states 51-100 are another, and so on.)\n",
    "\n",
    "Hence, 50 states would share the same feature and value estimate, and there would be 10 distinct features. The feature vector for each state is a one-hot feature vector of length 10, with a single one indicating the group for that state. (one-hot vector of length 10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "markdown",
     "checksum": "84dc88656a91900d7fb23820f5eabdaa",
     "grade": false,
     "grade_id": "cell-3bd24c4fbda046a9",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "source": [
    "## Section 2-1: Implement Useful Functions\n",
    "\n",
    "Before we implement the agent, we need to define a couple of useful helper functions.\n",
    "\n",
    "**Please note all random method calls should be called through random number generator. Also do not use random method calls unless specified. In the agent, only `agent_policy` requires random method calls.**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "markdown",
     "checksum": "030512a7cb2cbea62cec634b633aeba0",
     "grade": false,
     "grade_id": "cell-fbf1a1330b3b6c3b",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "source": [
    "## Section 2-1a: Selecting actions\n",
    "\n",
    "In this part we have implemented `agent_policy()` for you.\n",
    "\n",
    "This method is used in `agent_start()` and `agent_step()` to select appropriate action.\n",
    "Normally, the agent acts differently given state, but in this environment the agent chooses randomly to move either left or right with equal probability.\n",
    "\n",
    "Agent returns 0 for left, and 1 for right."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "code",
     "checksum": "03bb830eea52c7388af728e97748adfe",
     "grade": false,
     "grade_id": "cell-70fd3462c0981100",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "outputs": [],
   "source": [
    "# ---------------\n",
    "# Discussion Cell\n",
    "# ---------------\n",
    "\n",
    "def agent_policy(rand_generator, state):\n",
    "    \"\"\"\n",
    "    Given random number generator and state, returns an action according to the agent's policy.\n",
    "    \n",
    "    Args:\n",
    "        rand_generator: Random number generator\n",
    "\n",
    "    Returns:\n",
    "        chosen action [int]\n",
    "    \"\"\"\n",
    "    \n",
    "    # set chosen_action as 0 or 1 with equal probability\n",
    "    # state is unnecessary for this agent policy\n",
    "    chosen_action = rand_generator.choice([0,1])\n",
    "    \n",
    "    return chosen_action"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "markdown",
     "checksum": "0b11c104f3dc5013cfd73884bf7875eb",
     "grade": false,
     "grade_id": "cell-26e7a90b0b26c7a0",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "source": [
    "## Section 2-1b: Processing State Features with State Aggregation\n",
    "\n",
    "In this part you will implement `get_state_feature()`\n",
    "\n",
    "This method takes in a state and returns the aggregated feature (one-hot-vector) of that state.\n",
    "The feature vector size is determined by `num_groups`. Use `state` and `num_states_in_group` to determine which element in the feature vector is active.\n",
    "\n",
    "`get_state_feature()` is necessary whenever the agent receives a state and needs to convert it to a feature for learning. The features will thus be used in `agent_step()` and `agent_end()` when the agent updates its state values."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "deletable": false,
    "nbgrader": {
     "cell_type": "code",
     "checksum": "13d3aaef058d482e6f4e630d0975504a",
     "grade": false,
     "grade_id": "cell-0c3f9e12091de241",
     "locked": false,
     "schema_version": 3,
     "solution": true,
     "task": false
    }
   },
   "outputs": [],
   "source": [
    "# -----------\n",
    "# Graded Cell\n",
    "# -----------\n",
    "\n",
    "def get_state_feature(num_states_in_group, num_groups, state):\n",
    "    \"\"\"\n",
    "    Given state, return the feature of that state\n",
    "    \n",
    "    Args:\n",
    "        num_states_in_group [int]\n",
    "        num_groups [int] \n",
    "        state [int] : 1~500\n",
    "\n",
    "    Returns:\n",
    "        one_hot_vector [numpy array]\n",
    "    \"\"\"\n",
    "    \n",
    "    ### Generate state feature (2~4 lines)\n",
    "    # Create one_hot_vector with size of the num_groups, according to state\n",
    "    # For simplicity, assume num_states is always perfectly divisible by num_groups\n",
    "    # Note that states start from index 1, not 0!\n",
    "    \n",
    "    # Example:\n",
    "    # If num_states = 100, num_states_in_group = 20, num_groups = 5,\n",
    "    # one_hot_vector would be of size 5.\n",
    "    # For states 1~20, one_hot_vector would be: [1, 0, 0, 0, 0]\n",
    "    # \n",
    "    # one_hot_vector = ?\n",
    "    \n",
    "    # ----------------\n",
    "    # your code here\n",
    "    \n",
    "    one_hot_vector = np.zeros(num_groups)\n",
    "    one_hot_vector[(state - 1) // num_states_in_group] = 1   \n",
    "    \n",
    "    return one_hot_vector\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "markdown",
     "checksum": "da48819a930ca982643dd9fe891adfde",
     "grade": false,
     "grade_id": "cell-a9bca365b5e9faa6",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "source": [
    "Run the following code to verify your `get_state_feature()` function."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "code",
     "checksum": "222c94a1ddaa58d6318d97e182519542",
     "grade": true,
     "grade_id": "cell-f03905f4b2fe4cd2",
     "locked": true,
     "points": 10,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1st group: [1. 0. 0. 0. 0.]\n",
      "2nd group: [0. 1. 0. 0. 0.]\n",
      "3rd group: [0. 0. 1. 0. 0.]\n",
      "4th group: [0. 0. 0. 1. 0.]\n",
      "5th group: [0. 0. 0. 0. 1.]\n"
     ]
    }
   ],
   "source": [
    "# -----------\n",
    "# Tested Cell\n",
    "# -----------\n",
    "# The contents of the cell will be tested by the autograder.\n",
    "# If they do not pass here, they will not pass there.\n",
    "\n",
    "# Given that num_states = 10 and num_groups = 5, test get_state_feature()\n",
    "# There are states 1~10, and the state feature vector would be of size 5.\n",
    "# Only one element would be active for any state feature vector.\n",
    "\n",
    "# get_state_feature() should support various values of num_states, num_groups, not just this example\n",
    "# For simplicity, assume num_states will always be perfectly divisible by num_groups\n",
    "num_states = 10\n",
    "num_groups = 5\n",
    "num_states_in_group = int(num_states / num_groups)\n",
    "\n",
    "# Test 1st group, state = 1\n",
    "state = 1\n",
    "features = get_state_feature(num_states_in_group, num_groups, state)\n",
    "print(\"1st group: {}\".format(features))\n",
    "\n",
    "assert np.all(features == [1, 0, 0, 0, 0])\n",
    "\n",
    "# Test 2nd group, state = 3\n",
    "state = 3\n",
    "features = get_state_feature(num_states_in_group, num_groups, state)\n",
    "print(\"2nd group: {}\".format(features))\n",
    "\n",
    "assert np.all(features == [0, 1, 0, 0, 0])\n",
    "\n",
    "# Test 3rd group, state = 6\n",
    "state = 6\n",
    "features = get_state_feature(num_states_in_group, num_groups, state)\n",
    "print(\"3rd group: {}\".format(features))\n",
    "\n",
    "assert np.all(features == [0, 0, 1, 0, 0])\n",
    "\n",
    "# Test 4th group, state = 7\n",
    "state = 7\n",
    "features = get_state_feature(num_states_in_group, num_groups, state)\n",
    "print(\"4th group: {}\".format(features))\n",
    "\n",
    "assert np.all(features == [0, 0, 0, 1, 0])\n",
    "\n",
    "# Test 5th group, state = 10\n",
    "state = 10\n",
    "features = get_state_feature(num_states_in_group, num_groups, state)\n",
    "print(\"5th group: {}\".format(features))\n",
    "\n",
    "assert np.all(features == [0, 0, 0, 0, 1])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "markdown",
     "checksum": "c9bf6ee73b3ced770f28fb786acca746",
     "grade": false,
     "grade_id": "cell-88b7a7b538f47cca",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "source": [
    "## Section 2-2: Implement Agent Methods\n",
    "\n",
    "Now that we have implemented all the helper functions, let's create an agent. In this part, you will implement `agent_init()`, `agent_start()`, `agent_step()` and `agent_end()`. You will have to use `agent_policy()` that we implemented above. We will implement `agent_message()` later, when returning the learned state-values.\n",
    "\n",
    "To save computation time, we precompute features for all states beforehand in `agent_init()`. The pre-computed features are saved in `self.all_state_features` numpy array. Hence, you do not  need to call `get_state_feature()` every time in `agent_step()` and `agent_end()`.\n",
    "\n",
    "The shape of `self.all_state_features` numpy array is `(num_states, feature_size)`, with features of states from State 1-500. Note that index 0 stores features for State 1 (Features for State 0 does not exist). Use `self.all_state_features` to access each feature vector for a state.\n",
    "\n",
    "When saving state values in the agent, recall how the state values are represented with linear function approximation.\n",
    "\n",
    "**State Value Representation**: $\\hat{v}(s,\\mathbf{w}) = \\mathbf{w}\\cdot\\mathbf{x^T}$ where $\\mathbf{w}$ is a weight vector and $\\mathbf{x}$ is the feature vector of the state.\n",
    "\n",
    "\n",
    "When performing TD(0) updates with Linear Function Approximation, recall how we perform semi-gradient TD(0) updates using supervised learning.\n",
    "\n",
    "**semi-gradient TD(0) Weight Update Rule**: $\\mathbf{w_{t+1}} = \\mathbf{w_{t}} + \\alpha [R_{t+1} + \\gamma \\hat{v}(S_{t+1},\\mathbf{w}) - \\hat{v}(S_t,\\mathbf{w})] \\nabla \\hat{v}(S_t,\\mathbf{w})$"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "deletable": false,
    "nbgrader": {
     "cell_type": "code",
     "checksum": "e1f3a8172c159929cde4ace3e57b850a",
     "grade": false,
     "grade_id": "cell-5f5dd660fd400570",
     "locked": false,
     "schema_version": 3,
     "solution": true,
     "task": false
    }
   },
   "outputs": [],
   "source": [
    "# -----------\n",
    "# Graded Cell\n",
    "# -----------\n",
    "\n",
    "# Create TDAgent\n",
    "class TDAgent(BaseAgent):\n",
    "    def __init__(self):\n",
    "        self.num_states = None\n",
    "        self.num_groups = None\n",
    "        self.step_size = None\n",
    "        self.discount_factor = None\n",
    "        \n",
    "    def agent_init(self, agent_info={}):\n",
    "        \"\"\"Setup for the agent called when the experiment first starts.\n",
    "\n",
    "        Set parameters needed to setup the semi-gradient TD(0) state aggregation agent.\n",
    "\n",
    "        Assume agent_info dict contains:\n",
    "        {\n",
    "            num_states: 500 [int],\n",
    "            num_groups: int, \n",
    "            step_size: float, \n",
    "            discount_factor: float,\n",
    "            seed: int\n",
    "        }\n",
    "        \"\"\"\n",
    "\n",
    "        # set random seed for each run\n",
    "        self.rand_generator = np.random.RandomState(agent_info.get(\"seed\")) \n",
    "\n",
    "        # set class attributes\n",
    "        self.num_states = agent_info.get(\"num_states\")\n",
    "        self.num_groups = agent_info.get(\"num_groups\")\n",
    "        self.step_size = agent_info.get(\"step_size\")\n",
    "        self.discount_factor = agent_info.get(\"discount_factor\")\n",
    "\n",
    "        # pre-compute all observable features\n",
    "        num_states_in_group = int(self.num_states / self.num_groups)\n",
    "        self.all_state_features = np.array([get_state_feature(num_states_in_group, self.num_groups, state) for state in range(1, self.num_states + 1)])\n",
    "\n",
    "        # ----------------\n",
    "        # initialize all weights to zero using numpy array with correct size\n",
    "        # self.weights = ?\n",
    "        # your code here\n",
    "        \n",
    "        self.weights = np.zeros(self.num_groups)\n",
    "\n",
    "        self.last_state = None\n",
    "        self.last_action = None\n",
    "\n",
    "    def agent_start(self, state):\n",
    "        \"\"\"The first method called when the experiment starts, called after\n",
    "        the environment starts.\n",
    "        Args:\n",
    "            state [int]: the state from the\n",
    "                environment's evn_start function.\n",
    "        Returns:\n",
    "            self.last_action [int] : The first action the agent takes.\n",
    "        \"\"\"\n",
    "\n",
    "        # ----------------\n",
    "        ### select action given state (using agent_policy), and save current state and action\n",
    "        # Use self.rand_generator for agent_policy\n",
    "        # \n",
    "        # self.last_state = ?\n",
    "        # self.last_action = ?\n",
    "        # your code here\n",
    "        \n",
    "        self.last_state = state\n",
    "        self.last_action = agent_policy(self.rand_generator, state)\n",
    "\n",
    "        return self.last_action\n",
    "\n",
    "    def agent_step(self, reward, state):\n",
    "        \"\"\"A step taken by the agent.\n",
    "        Args:\n",
    "            reward [float]: the reward received for taking the last action taken\n",
    "            state [int]: the state from the environment's step, where the agent ended up after the last step\n",
    "        Returns:\n",
    "            self.last_action [int] : The action the agent is taking.\n",
    "        \"\"\"\n",
    "        \n",
    "        # get relevant feature\n",
    "        current_state_feature = self.all_state_features[state-1] \n",
    "        last_state_feature = self.all_state_features[self.last_state-1] \n",
    "        \n",
    "        ### update weights and select action\n",
    "        # (Hint: np.dot method is useful!)\n",
    "        #\n",
    "        # Update weights:\n",
    "        #     use self.weights, current_state_feature, and last_state_feature\n",
    "        #\n",
    "        # Select action:\n",
    "        #     use self.rand_generator for agent_policy\n",
    "        #\n",
    "        # Current state and selected action should be saved to self.last_state and self.last_action at the end\n",
    "        #\n",
    "        # self.weights = ?\n",
    "        # self.last_state = ?\n",
    "        # self.last_action = ?\n",
    "\n",
    "        # ----------------\n",
    "        # your code here\n",
    "       \n",
    "        self.weights += self.step_size * (\\\n",
    "                            reward + self.discount_factor *\\\n",
    "                                self.weights@current_state_feature.T -\\\n",
    "                                self.weights@last_state_feature.T) *\\\n",
    "                                last_state_feature # delta V(s,w) is e.g [1,0,0,0,0]\n",
    "        self.last_state = state\n",
    "        self.last_action = agent_policy(self.rand_generator, state)\n",
    "        \n",
    "    \n",
    "        return self.last_action\n",
    "\n",
    "    def agent_end(self, reward):\n",
    "        \"\"\"Run when the agent terminates.\n",
    "        Args:\n",
    "            reward (float): the reward the agent received for entering the\n",
    "                terminal state.\n",
    "        \"\"\"\n",
    "\n",
    "        # get relevant feature\n",
    "        last_state_feature = self.all_state_features[self.last_state-1]\n",
    "        \n",
    "        ### update weights\n",
    "        # Update weights using self.weights and last_state_feature\n",
    "        # (Hint: np.dot method is useful!)\n",
    "        # \n",
    "        # Note that here you don't need to choose action since the agent has reached a terminal state\n",
    "        # Therefore you should not update self.last_state and self.last_action\n",
    "        # \n",
    "        # self.weights = ?\n",
    "        \n",
    "        # ----------------\n",
    "        # your code here\n",
    "        \n",
    "        self.weights += self.step_size * (\\\n",
    "                        reward - self.weights@last_state_feature.T) * last_state_feature\n",
    "        \n",
    "        return\n",
    "        \n",
    "    def agent_message(self, message):\n",
    "        # We will implement this method later\n",
    "        raise NotImplementedError\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "markdown",
     "checksum": "06b8089562061cb16a442f14afa90bff",
     "grade": false,
     "grade_id": "cell-5cde0a383d7244f6",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "source": [
    "\n",
    "Run the following code to verify `agent_init()`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "code",
     "checksum": "bb1d63d54f56543a7b3eb2866b15b7b6",
     "grade": true,
     "grade_id": "cell-51482f87693745b0",
     "locked": true,
     "points": 5,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "num_states: 500\n",
      "num_groups: 10\n",
      "step_size: 0.1\n",
      "discount_factor: 1.0\n",
      "weights shape: (10,)\n",
      "weights init. value: [0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n"
     ]
    }
   ],
   "source": [
    "# -----------\n",
    "# Tested Cell\n",
    "# -----------\n",
    "# The contents of the cell will be tested by the autograder.\n",
    "# If they do not pass here, they will not pass there.\n",
    "\n",
    "agent_info = {\n",
    "    \"num_states\": 500,\n",
    "    \"num_groups\": 10,\n",
    "    \"step_size\": 0.1,\n",
    "    \"discount_factor\": 1.0,\n",
    "    \"seed\": 1,\n",
    "}\n",
    "\n",
    "agent = TDAgent()\n",
    "agent.agent_init(agent_info)\n",
    "\n",
    "assert np.all(agent.weights == 0)\n",
    "assert agent.weights.shape == (10,)\n",
    "\n",
    "# check attributes\n",
    "print(\"num_states: {}\".format(agent.num_states))\n",
    "print(\"num_groups: {}\".format(agent.num_groups))\n",
    "print(\"step_size: {}\".format(agent.step_size))\n",
    "print(\"discount_factor: {}\".format(agent.discount_factor))\n",
    "\n",
    "print(\"weights shape: {}\".format(agent.weights.shape))\n",
    "print(\"weights init. value: {}\".format(agent.weights))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "markdown",
     "checksum": "b8e3cd706b80ba48ea1b029dc6edf271",
     "grade": false,
     "grade_id": "cell-a35feea9e9c0d939",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "source": [
    "Run the following code to verify `agent_start()`.\n",
    "Although there is randomness due to `rand_generator.choice()` in `agent_policy()`, we control the seed so your output should match the expected output. \n",
    "\n",
    "Make sure `rand_generator.choice()` is called only once per `agent_policy()` call."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "code",
     "checksum": "8be0a3ee9c6a85cd19446cc7484fc483",
     "grade": true,
     "grade_id": "cell-b33fdd3efc9215da",
     "locked": true,
     "points": 5,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Agent state: 250\n",
      "Agent selected action: 1\n"
     ]
    }
   ],
   "source": [
    "# -----------\n",
    "# Tested Cell\n",
    "# -----------\n",
    "# The contents of the cell will be tested by the autograder.\n",
    "# If they do not pass here, they will not pass there.\n",
    "\n",
    "agent_info = {\n",
    "    \"num_states\": 500,\n",
    "    \"num_groups\": 10,\n",
    "    \"step_size\": 0.1,\n",
    "    \"discount_factor\": 1.0,\n",
    "    \"seed\": 1,\n",
    "}\n",
    "\n",
    "# Suppose state = 250\n",
    "state = 250\n",
    "\n",
    "agent = TDAgent()\n",
    "agent.agent_init(agent_info)\n",
    "action = agent.agent_start(state)\n",
    "\n",
    "assert action == 1\n",
    "assert agent.last_state == 250\n",
    "assert agent.last_action == 1\n",
    "\n",
    "print(\"Agent state: {}\".format(agent.last_state))\n",
    "print(\"Agent selected action: {}\".format(agent.last_action))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "markdown",
     "checksum": "ee08e5d2d9db855447d517f4b7ea96d7",
     "grade": false,
     "grade_id": "cell-b3e1d1fcdeec63fb",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "source": [
    "Run the following code to verify `agent_step()`\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "code",
     "checksum": "3c3420b0c2fb9120a9bee034ad9f6987",
     "grade": true,
     "grade_id": "cell-f06dfcb8e4b952c9",
     "locked": true,
     "points": 20,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Updated weights: [-0.26  0.5   1.   -0.5   1.5  -0.5   1.5   0.   -0.5  -1.  ]\n",
      "last state: 120\n",
      "last action: 1\n"
     ]
    }
   ],
   "source": [
    "# -----------\n",
    "# Tested Cell\n",
    "# -----------\n",
    "# The contents of the cell will be tested by the autograder.\n",
    "# If they do not pass here, they will not pass there.\n",
    "\n",
    "agent_info = {\n",
    "    \"num_states\": 500,\n",
    "    \"num_groups\": 10,\n",
    "    \"step_size\": 0.1,\n",
    "    \"discount_factor\": 0.9,\n",
    "    \"seed\": 1,\n",
    "}\n",
    "\n",
    "agent = TDAgent()\n",
    "agent.agent_init(agent_info)\n",
    "\n",
    "# Initializing the weights to arbitrary values to verify the correctness of weight update\n",
    "agent.weights = np.array([-1.5, 0.5, 1., -0.5, 1.5, -0.5, 1.5, 0.0, -0.5, -1.0])\n",
    "\n",
    "# Assume the agent started at State 50\n",
    "start_state = 50\n",
    "action = agent.agent_start(start_state)\n",
    "\n",
    "assert action == 1\n",
    "\n",
    "# Assume the reward was 10.0 and the next state observed was State 120\n",
    "reward = 10.0\n",
    "next_state = 120\n",
    "action = agent.agent_step(reward, next_state)\n",
    "\n",
    "assert action == 1\n",
    "\n",
    "print(\"Updated weights: {}\".format(agent.weights))\n",
    "assert np.allclose(agent.weights, [-0.26, 0.5, 1., -0.5, 1.5, -0.5, 1.5, 0., -0.5, -1.])\n",
    "\n",
    "assert agent.last_state == 120\n",
    "assert agent.last_action == 1\n",
    "\n",
    "print(\"last state: {}\".format(agent.last_state))\n",
    "print(\"last action: {}\".format(agent.last_action))\n",
    "\n",
    "# let's do another\n",
    "reward = -22\n",
    "next_state = 222\n",
    "action = agent.agent_step(reward, next_state)\n",
    "\n",
    "assert action == 0\n",
    "\n",
    "assert np.allclose(agent.weights, [-0.26, 0.5, -1.165, -0.5, 1.5, -0.5, 1.5, 0, -0.5, -1])\n",
    "assert agent.last_state == 222\n",
    "assert agent.last_action == 0"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "markdown",
     "checksum": "a5d3775348b0d2f66e8092672f419eae",
     "grade": false,
     "grade_id": "cell-b51199488eba4b22",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "source": [
    "Run the following code to verify `agent_end()`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "code",
     "checksum": "4d298a9af62ff0f53efd137439b2b45a",
     "grade": true,
     "grade_id": "cell-3449ab4cb6f74b31",
     "locked": true,
     "points": 10,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Updated weights: [-0.35  0.5   1.   -0.5   1.5  -0.5   1.5   0.   -0.5  -1.  ]\n"
     ]
    }
   ],
   "source": [
    "# -----------\n",
    "# Tested Cell\n",
    "# -----------\n",
    "# The contents of the cell will be tested by the autograder.\n",
    "# If they do not pass here, they will not pass there.\n",
    "\n",
    "agent_info = {\n",
    "    \"num_states\": 500,\n",
    "    \"num_groups\": 10,\n",
    "    \"step_size\": 0.1,\n",
    "    \"discount_factor\": 0.9,\n",
    "    \"seed\": 1,\n",
    "}\n",
    "\n",
    "agent = TDAgent()\n",
    "agent.agent_init(agent_info)\n",
    "\n",
    "# Initializing the weights to arbitrary values to verify the correctness of weight update\n",
    "agent.weights = np.array([-1.5, 0.5, 1., -0.5, 1.5, -0.5, 1.5, 0.0, -0.5, -1.0])\n",
    "\n",
    "# Assume the agent started at State 50\n",
    "start_state = 50\n",
    "action = agent.agent_start(start_state)\n",
    "\n",
    "assert action == 1\n",
    "\n",
    "# Assume the reward was 10.0 and reached the terminal state\n",
    "agent.agent_end(10.0)\n",
    "print(\"Updated weights: {}\".format(agent.weights))\n",
    "\n",
    "assert np.allclose(agent.weights, [-0.35, 0.5, 1., -0.5, 1.5, -0.5, 1.5, 0., -0.5, -1.])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "markdown",
     "checksum": "7351d74028871274b8fc9c69f1c0502e",
     "grade": false,
     "grade_id": "cell-92f56709f6a305ee",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "source": [
    "**Expected output**: (Note only the 1st element was changed, and the result is different from `agent_step()` )\n",
    "    \n",
    "    Initial weights: [-1.5  0.5  1.  -0.5  1.5 -0.5  1.5  0.  -0.5 -1. ]\n",
    "    Updated weights: [-0.35  0.5   1.   -0.5   1.5  -0.5   1.5   0.   -0.5  -1.  ]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "markdown",
     "checksum": "a14cc0415ad4db72bd3b9d156b45cc38",
     "grade": false,
     "grade_id": "cell-31513fa5e20f67da",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "source": [
    "## Section 2-3: Returning Learned State Values\n",
    "\n",
    "You are almost done! Now let's implement a code block in `agent_message()` that returns the learned state values.\n",
    "\n",
    "The method `agent_message()` will return the learned state_value array when `message == 'get state value'`.\n",
    "\n",
    "**Hint**: Think about how state values are represented with linear function approximation. `state_value` array will be a 1D array with length equal to the number of states."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "deletable": false,
    "nbgrader": {
     "cell_type": "code",
     "checksum": "eb78c449ca2c4313477fa2279290046b",
     "grade": false,
     "grade_id": "cell-71fa8b08ea15ff38",
     "locked": false,
     "schema_version": 3,
     "solution": true,
     "task": false
    }
   },
   "outputs": [],
   "source": [
    "%%add_to TDAgent\n",
    "\n",
    "# -----------\n",
    "# Graded Cell\n",
    "# -----------\n",
    "\n",
    "def agent_message(self, message):\n",
    "    if message == 'get state value':\n",
    "        \n",
    "        ### return state_value\n",
    "        # Use self.all_state_features and self.weights to return the vector of all state values\n",
    "        # Hint: Use np.dot()\n",
    "        #\n",
    "        # state_value = ?\n",
    "        state_value =  self.weights @ self.all_state_features.T\n",
    "        # your code here\n",
    "        \n",
    "        \n",
    "        return state_value"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "markdown",
     "checksum": "00b2355d8247a3cf39bf63af25343cad",
     "grade": false,
     "grade_id": "cell-63400d109704d1cd",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "source": [
    "Run the following code to verify `get_state_val()`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "code",
     "checksum": "cbcd8f8f3120bbe709a8b53fb4d57d15",
     "grade": true,
     "grade_id": "cell-f8d18149b05b7c49",
     "locked": true,
     "points": 5,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "State value shape: (20,)\n",
      "Initial State value for all states: [0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n"
     ]
    }
   ],
   "source": [
    "# -----------\n",
    "# Tested Cell\n",
    "# -----------\n",
    "# The contents of the cell will be tested by the autograder.\n",
    "# If they do not pass here, they will not pass there.\n",
    "\n",
    "agent_info = {\n",
    "    \"num_states\": 20,\n",
    "    \"num_groups\": 5,\n",
    "    \"step_size\": 0.1,\n",
    "    \"discount_factor\": 1.0,\n",
    "}\n",
    "\n",
    "agent = TDAgent()\n",
    "agent.agent_init(agent_info)\n",
    "test_state_val = agent.agent_message('get state value')\n",
    "\n",
    "assert test_state_val.shape == (20,)\n",
    "assert np.all(test_state_val == 0)\n",
    "\n",
    "print(\"State value shape: {}\".format(test_state_val.shape))\n",
    "print(\"Initial State value for all states: {}\".format(test_state_val))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "markdown",
     "checksum": "0674ef3c418b263db0c0d9bae7675292",
     "grade": false,
     "grade_id": "cell-3b92a95f34b9560e",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "source": [
    "**Expected Output**:\n",
    "\n",
    "    State value shape: (20,)\n",
    "    Initial State value for all states: [0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "markdown",
     "checksum": "e9093617622567a25678f5456cfedc5e",
     "grade": false,
     "grade_id": "cell-9a752ea33a9a9c10",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "source": [
    "## Section 3: Run Experiment\n",
    "\n",
    "Now that we've implemented all the components of environment and agent, let's run an experiment! We will plot two things: (1) the learned state value function and compare it against the true state values, and (2) a learning curve depicting the error in the learned value estimates over episodes. For the learning curve, what should we plot to see if the agent is learning well?"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "markdown",
     "checksum": "a045679989ba171d36ca6a2b7366da33",
     "grade": false,
     "grade_id": "cell-8286e23b6eeb2ebe",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "source": [
    "## Section 3-1: Prediction Objective (Root Mean Squared Value Error) \n",
    "\n",
    "Recall that the Prediction Objective in function approximation is Mean Squared Value Error $\\overline{VE}(\\mathbf{w}) \\doteq \\sum\\limits_{s \\in \\mathcal{S}}\\mu(s)[v_\\pi(s)-\\hat{v}(s,\\mathbf{w})]^2$\n",
    "\n",
    "We will use the square root of this measure, the root $\\overline{VE}$ to give a rough measure of how much the learned values differ from the true values.\n",
    "\n",
    "`calc RMSVE()` computes the Root Mean Squared Value Error given learned state value $\\hat{v}(s, \\mathbf{w})$.\n",
    "We provide you with true state value $v_\\pi(s)$ and state distribution $\\mu(s)$\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "code",
     "checksum": "2fa55c90fc3679615a86df3dac8c8522",
     "grade": false,
     "grade_id": "cell-f974bf03a35b3975",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "outputs": [],
   "source": [
    "# ---------------\n",
    "# Discussion Cell\n",
    "# ---------------\n",
    "\n",
    "# Here we provide you with the true state value and state distribution\n",
    "true_state_val = np.load('data/true_V.npy')    \n",
    "state_distribution = np.load('data/state_distribution.npy')\n",
    "\n",
    "def calc_RMSVE(learned_state_val):\n",
    "    assert(len(true_state_val) == len(learned_state_val) == len(state_distribution))\n",
    "    MSVE = np.sum(np.multiply(state_distribution, np.square(true_state_val - learned_state_val)))\n",
    "    RMSVE = np.sqrt(MSVE)\n",
    "    return RMSVE"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "markdown",
     "checksum": "6728c5c1a540545b395a0577cd097459",
     "grade": false,
     "grade_id": "cell-bc8f5ef76e2ce172",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "source": [
    "## Section 3-2a: Run Experiment with 10-State Aggregation\n",
    "\n",
    "We have provided you the experiment/plot code in the cell below."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "code",
     "checksum": "b1ae63f2204e736a41d57fc2674260b7",
     "grade": false,
     "grade_id": "cell-2aec1de29d4beb79",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "outputs": [],
   "source": [
    "# ---------------\n",
    "# Discussion Cell\n",
    "# ---------------\n",
    "\n",
    "import os\n",
    "\n",
    "# Define function to run experiment\n",
    "def run_experiment(environment, agent, environment_parameters, agent_parameters, experiment_parameters):\n",
    "\n",
    "    rl_glue = RLGlue(environment, agent)\n",
    "    \n",
    "    # Sweep Agent parameters\n",
    "    for num_agg_states in agent_parameters[\"num_groups\"]:\n",
    "        for step_size in agent_parameters[\"step_size\"]:\n",
    "            \n",
    "            # save rmsve at the end of each evaluation episode\n",
    "            # size: num_episode / episode_eval_frequency + 1 (includes evaluation at the beginning of training)\n",
    "            agent_rmsve = np.zeros(int(experiment_parameters[\"num_episodes\"]/experiment_parameters[\"episode_eval_frequency\"]) + 1)\n",
    "            \n",
    "            # save learned state value at the end of each run\n",
    "            agent_state_val = np.zeros(environment_parameters[\"num_states\"])\n",
    "\n",
    "            env_info = {\"num_states\": environment_parameters[\"num_states\"],\n",
    "                        \"start_state\": environment_parameters[\"start_state\"],\n",
    "                        \"left_terminal_state\": environment_parameters[\"left_terminal_state\"],\n",
    "                        \"right_terminal_state\": environment_parameters[\"right_terminal_state\"]}\n",
    "\n",
    "            agent_info = {\"num_states\": environment_parameters[\"num_states\"],\n",
    "                          \"num_groups\": num_agg_states,\n",
    "                          \"step_size\": step_size,\n",
    "                          \"discount_factor\": environment_parameters[\"discount_factor\"]}\n",
    "\n",
    "            print('Setting - num. agg. states: {}, step_size: {}'.format(num_agg_states, step_size))\n",
    "            os.system('sleep 0.2')\n",
    "            \n",
    "            # one agent setting\n",
    "            for run in tqdm(range(1, experiment_parameters[\"num_runs\"]+1)):\n",
    "                env_info[\"seed\"] = run\n",
    "                agent_info[\"seed\"] = run\n",
    "                rl_glue.rl_init(agent_info, env_info)\n",
    "                \n",
    "                # Compute initial RMSVE before training\n",
    "                current_V = rl_glue.rl_agent_message(\"get state value\")\n",
    "                agent_rmsve[0] += calc_RMSVE(current_V)\n",
    "                    \n",
    "                for episode in range(1, experiment_parameters[\"num_episodes\"]+1):\n",
    "                    # run episode\n",
    "                    rl_glue.rl_episode(0) # no step limit\n",
    "                    \n",
    "                    if episode % experiment_parameters[\"episode_eval_frequency\"] == 0:\n",
    "                        current_V = rl_glue.rl_agent_message(\"get state value\")\n",
    "                        agent_rmsve[int(episode/experiment_parameters[\"episode_eval_frequency\"])] += calc_RMSVE(current_V)\n",
    "                        \n",
    "                # store only one run of state value\n",
    "                if run == 50:\n",
    "                    agent_state_val = rl_glue.rl_agent_message(\"get state value\")\n",
    "            \n",
    "            # rmsve averaged over runs\n",
    "            agent_rmsve /= experiment_parameters[\"num_runs\"]\n",
    "            \n",
    "            save_name = \"{}_agg_states_{}_step_size_{}\".format('TD_agent', num_agg_states, step_size).replace('.','')\n",
    "            \n",
    "            if not os.path.exists('results'):\n",
    "                os.makedirs('results')\n",
    "    \n",
    "            # save avg. state value\n",
    "            np.save(\"results/V_{}\".format(save_name), agent_state_val)\n",
    "\n",
    "            # save avg. rmsve\n",
    "            np.save(\"results/RMSVE_{}\".format(save_name), agent_rmsve)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "markdown",
     "checksum": "2bdd48cb4012e928bb85c3127836d9f6",
     "grade": false,
     "grade_id": "cell-1c5474a596986e5c",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "source": [
    "\n",
    "We will first test our implementation using state aggregation with resolution of 10, with three different step sizes: {0.01, 0.05, 0.1}.\n",
    "\n",
    "Note that running the experiment cell below will take **_approximately 5 min_**.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "code",
     "checksum": "621dbc134b486d99d34f8e63fd3b2468",
     "grade": false,
     "grade_id": "cell-0c0267acbc817952",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Setting - num. agg. states: 10, step_size: 0.01\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 50/50 [02:00<00:00,  2.41s/it]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Setting - num. agg. states: 10, step_size: 0.05\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 50/50 [01:56<00:00,  2.33s/it]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Setting - num. agg. states: 10, step_size: 0.1\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 50/50 [01:57<00:00,  2.35s/it]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x360 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# ---------------\n",
    "# Discussion Cell\n",
    "# ---------------\n",
    "\n",
    "#### Run Experiment\n",
    "\n",
    "# Experiment parameters\n",
    "experiment_parameters = {\n",
    "    \"num_runs\" : 50,\n",
    "    \"num_episodes\" : 2000,\n",
    "    \"episode_eval_frequency\" : 10 # evaluate every 10 episodes\n",
    "}\n",
    "\n",
    "# Environment parameters\n",
    "environment_parameters = {\n",
    "    \"num_states\" : 500, \n",
    "    \"start_state\" : 250,\n",
    "    \"left_terminal_state\" : 0,\n",
    "    \"right_terminal_state\" : 501, \n",
    "    \"discount_factor\" : 1.0\n",
    "}\n",
    "\n",
    "# Agent parameters\n",
    "# Each element is an array because we will be later sweeping over multiple values\n",
    "agent_parameters = {\n",
    "    \"num_groups\": [10],\n",
    "    \"step_size\": [0.01, 0.05, 0.1]\n",
    "}\n",
    "\n",
    "current_env = RandomWalkEnvironment\n",
    "current_agent = TDAgent\n",
    "\n",
    "run_experiment(current_env, current_agent, environment_parameters, agent_parameters, experiment_parameters)\n",
    "plot_script.plot_result(agent_parameters, 'results')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "markdown",
     "checksum": "69dca4f482f095799ebce9bb62ab0b9e",
     "grade": false,
     "grade_id": "cell-4042482c3a7ca849",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "source": [
    "Is the learned state value plot with step-size=0.01 similar to Figure 9.2 (p.208) in Sutton and Barto?\n",
    "\n",
    "(Note that our environment has less states: 500 states and we have done 2000 episodes, and averaged the performance over 50 runs)\n",
    "\n",
    "Look at  the plot of the learning curve. Does RMSVE decrease over time?\n",
    "\n",
    "Would it be possible to reduce RMSVE to 0?\n",
    "\n",
    "You should see the RMSVE decrease over time, but the error seems to plateau. It is impossible to reduce RMSVE to 0, because of function approximation (and we do not decay the step-size parameter to zero). With function approximation, the agent has limited resources and has to trade-off the accuracy of one state for another state."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "markdown",
     "checksum": "dd21252c8db618346960f284b389bb75",
     "grade": false,
     "grade_id": "cell-32c4ce7f4fb916e6",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "source": [
    "Run the following code to verify your experimental result."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "code",
     "checksum": "95d8d26ff88c94ae7ece8a50e3e7fd42",
     "grade": true,
     "grade_id": "cell-f515287976239cff",
     "locked": true,
     "points": 10,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Your experiment results are correct!\n"
     ]
    }
   ],
   "source": [
    "# -----------\n",
    "# Graded Cell\n",
    "# -----------\n",
    "\n",
    "agent_parameters = {\n",
    "    \"num_groups\": [10],\n",
    "    \"step_size\": [0.01, 0.05, 0.1]\n",
    "}\n",
    "\n",
    "all_correct = True\n",
    "for num_agg_states in agent_parameters[\"num_groups\"]:\n",
    "    for step_size in agent_parameters[\"step_size\"]:\n",
    "        filename = 'RMSVE_TD_agent_agg_states_{}_step_size_{}'.format(num_agg_states, step_size).replace('.','')\n",
    "        agent_RMSVE = np.load('results/{}.npy'.format(filename))\n",
    "        correct_RMSVE = np.load('correct_npy/{}.npy'.format(filename))\n",
    "\n",
    "        if not np.allclose(agent_RMSVE, correct_RMSVE):\n",
    "            all_correct=False\n",
    "\n",
    "if all_correct:\n",
    "    print(\"Your experiment results are correct!\")\n",
    "else:\n",
    "    print(\"Your experiment results does not match with ours. Please check if you have implemented all methods correctly.\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "markdown",
     "checksum": "1fde88cde6d01cbe2a1e8f3455857f17",
     "grade": false,
     "grade_id": "cell-c8ca7e4bb86e6b7c",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "source": [
    "## Section 3-2b: Run Experiment with Different State Aggregation Resolution and Step-Size\n",
    "\n",
    "In this section, we will run some more experiments to see how different parameter settings affect the results!\n",
    "\n",
    "In particular, we will test several values of `num_groups` and `step_size`. Parameter sweeps although necessary, can take lots of time. So now that you have verified your experiment result, here we show you the results of the parameter sweeps that you would see when running the sweeps yourself.\n",
    "\n",
    "We tested several different values of `num_groups`: {10, 100, 500}, and `step-size`: {0.01, 0.05, 0.1}. As before, we performed 2000 episodes per run, and averaged the results over 50 runs for each setting.\n",
    "\n",
    "Run the cell below to display the sweep results.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "code",
     "checksum": "f18c525408e2b34c54fcbe19239cca03",
     "grade": false,
     "grade_id": "cell-533a03fa5b239a32",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x360 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABDAAAAF2CAYAAACYrWjAAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+j8jraAAAgAElEQVR4nOzdd3wVVfr48c9DQKlSFYEg6FKUFEIPPRgBKQKigIiA0kRxV0WluL8IuKyyC7s2/C4WBEQFBARRUIoSigVERVkRKUuQEhVICMQkkPL8/pi5NzfJvSFIIEGf9+t1X7kzc+acMyUz554554yoKsYYY4wxxhhjjDHFWYmizoAxxhhjjDHGGGPM2VgFhjHGGGOMMcYYY4o9q8AwxhhjjDHGGGNMsWcVGMYYY4wxxhhjjCn2rALDGGOMMcYYY4wxxZ5VYBhjjDHGGGOMMabYswoMY4wxlzQRuVJEnhORLSJyWkTU/TwQIHxZEZkiIrvd8EdFZLGINPITNkhEHhaRHSKSKiKJIvKBiLQ5xzyGi8jbIvI/N83jIvK9m+6tucI+JCKTReShc9sTedKMcOOZLCIR5xNXAdJ62me/q4jcdyHTu1SJSB+fY1Ip17K6PvtvbhFl0RhjjCnWShZ1BowxxpjzVAv4S0ECikhJ4EOgvc/sasDtwM0iEqWqX/osmw8M9JkuDdwM3CQivVV1VQHSbAd8DJTymV3F/VwPJAHLfJY9BNQBDgDPFmS7AogAJrnf44Dt5xFXQCIiwJ25Zg8G/nMh0rvE9QGGut/nAieKLivGGGPMpcdaYBhjjLnUnQCeAe4AZp0l7H1kV168SXblRSZQHnjZE1BEbiG78uJjoAbQEfgV5wHAqyJyWQHyNwGn8iIL5wdsOZzKi1bAk8D/ChBHcdYBuCbXvNYi8qeiyExuIlK6qPNQEKoap6rifu4u6vwYY4wxxZFVYBhjjLmkuT/8xqrqIuDnswS/2+f7BFU9rqpLgQ3uvKYiEuYn7CRV/UlVNwKL3Hk1gK4FyGJ99+8pYI2qpqhqoqpuVdVJqvoUgIhEiYjitL4AqOPTpSDODRMhIu+IyF4ROSki6SLykzuvuSdBEYkF5vjkYY5PXHf7hLtTRDaKSJLbtWW3iPxdRMoWYLs8Bvt8f83n+6DcAUWkgoi84nbFOSkii0SklU/eYnOF7y4i34hImoj8ICKDRWSuT/gon7DeONyuGttF5AxOBRIiUkJExojIFyKS7HYJ2iEij7otc3zTreF2+Tnldvd5SURu8dfFQ0QGiMhaETkoIinufvyfiMwSkeq++SO79QXAfp/46ubXhURE2ovICnG6O3mO+UIRCc8VznfftBGRN9x9fVxElojI1QGOoTHGGHNJsAoMY4wxfwhuawnPD76TqnrIZ/F3Pt9b5Pqbe7m/sPk56P6tCOx2f9gOFZFrC7BubtcDtwJ/AirgtASp7s7bICI3FDQiEXkBpxVKe+AK4DKcypbH3bjKFCCOy4Hb3MmjwGPAGXf6rlxhBXgHGAFUcvPf353nL+4oYAXOMbscaAC8DnQ+S7bCgaVAY9xuOyJSwk1nJtAcpxVMaSAUmA4sc/PnabGxFuiH0yqnCjCKwK17OgE3AcFAGZz9eC1wL85+LEgrnYBE5C4gFrgFp8WQ55gPALb6VuLkshKnEqmSuw23AW+cT16MMcaYomYVGMYYY/4oqpI99lNSrmW+01e5f6sHWO4vbH6eBdT9Hozzw3Yu8D8R+VTcATZVNVZVBWfsC4ADPl0K6rrzvsJp9VED50f9FTjdYgDKunGjqlHAPT55uMcnrrkiEgl4BjmdC1ztrv+YO6+5T7z5uQXnBzLAclVNANa50/VFpJVP2JvcD8B+oBFQk8BdaP4OBLnfJ+FUAA1018lPZZyKmZpu3ubiVJT0dpc/jfOD/gqyxxjpiVMJBE6LkhD3+zagNk7lyckA6b2F0x2oGk6FSXWyW780BLoDuMd2ns961/ockzh/EYtIOeAFnPJahpvHK4DRbpDLgZcC5Gs/TkVXA+AXd160iNQIEN4YY4wp9qwCwxhjjAG5QGFR1feBaGA9zlgbvloD74tI+QJG95Mb18c4Y3+cJOdgmQ0LGM8tPt/vduNNwWmN4NGlAPH4dh9Zmusv5GyFEe3z/XlV/V5V43HGAcnB7cIS6U4mAH9X1ZOquhD49Cx5SgJGq2q8qia5lQO+2zvRjfMkzoCpHp7t9c3n31X1kKruAf4dIL144M/A1zj78GdyVh4V9Jj405bsCqJVqrpcVU+p6ktkD8raQETq+Vn3CVX9n5v3TT7z6/gJa4wxxlwSrALDGGPMH8VxnKfYkP2j0OMKn++ep9W+42lUyi+sO+6C5vrU9QRS1fWqeiPOU/oeON0R0t3FtXAqMgribWAccANOd4Xcztrtw1WQliNV81soIlVw3sgCkAr8IiKhwD6fYHf4jC9RzWf+jwG+e1Qhu4xyWFUzzxLe1w+qmpJr3rlsb4HzKSIVgc04FTW1yfmmGY+CHhN/rswn/QM+3/1t3w8+33/1+X5JDGpqjDHG+GMVGMYYY/4QVPUM8K07WUFEgn0Wh/h8/yLXX3C6O+QXNiAR8VZ4qOoJVV2lqvfhdG3wqOKb1QDxVCZ70NCf3XwEkT2uR25+43H94vN9kE9XBu8HaJnP+uCMweAZ36EMTveWHTjjNXhUI7uS45jP/Fo+32v7iTsB560tADXcMSzyC+8r1c883+1tG2B7+/+GfHYiu/LgI6CGG1eg1/rmd0z8OerzPfebXnynfyGvdJ/v55quMcYYUyxZBYYxxphLmvt2iWoiUg1nHAePcj7zPeb6fJ8mIlVF5Hac16MCfKWqO/yEnSIi1UWkI84Pd3C6DqwGZ8wJPz+K49xwy923QfQUkStFpJTbUqGDT/zf+3w/7v6tJiK+P6AzyP4hmoHTBaIa8LcAu+a4z/fQXG/aeN/n+1QRaSsipUWktoh0E5G38PMWkVwGn2W5h6cbyUc+8x4QkfrueAyTcq/gtqD43J2sBjwqzhtM7gDaFDBdX77b+7yINHa39zoR6Ssi75N9PHzzOV5EarpdNB7xE2+Gz/c04FcRCcHpUuKP7zFp7Bk4NB+fAInu924i0ktEyovISKCJO/8HVd17lniMMcaY3wWrwDDGGHOpuwbnSbXnLRge03zme/yH7PEABuE8bV+M05IhGedtEwCo6nvAAnfyRpxxImJx3mCRAYx0W3WczWVuWu/hPCk/g9NSwTM2wruq+q1PeM8P93LAIc9rNVX1FNk/rmvhvN3kZ3K2DvH1NdlvBHkESPd0bVHVz8geO+NanG4QqTjdFFbhDJZZkgBE5Dqyu70cAYJytWaoSHZLiF5uK5R1ZA/w2QDY7a77J5+ofVsK/JXsMUP+gVNhswDnOPgLn59FOG/lAGiGM35EKk53l6U43Xo8lQnzyX7TTFvgMLDH3abc6X5C9vnVw83jf/PJx+c+35cDWeK+ItcfVf0VpzIkC6d7yrs4r+N92Q1ymuwBPY0xxpjfPavAMMYY84ehqhk4XRr+BuzF+YF/HFgCtFLVL3OtMgQYi/Oj9DTOAJEfAh1VdSUFEwM8h/NGi3icpv0pOD+iHye7RYfHZGAhOStePO7C+TGe6OblDT/re7b1sJv/nW7ecy+/341vgxtXOnAIZ7DRccAH+WyT7+Ccb6pqlu9CVT2J82MbnO4lt6mq4rzK81U3vVM4lUejfFY97hNHLM6bQ3bgHKfdwFCcrip5wufHzV9vnDevbMGprDqNM47Eh+78r9ywaTival2CM3ZEIk6FwV9zp6uqiUA3nAqgFJwKmck4lWf+LMV5u8qP5B3QNVDe3wSicFqReMZx+RlnPJSW7n4yxhhj/hDEKU8YY4wxxlxYItIUOKaqP7rTlXFeOep5xel9qjrLXVYKZ4yJ9aqa7s67GaflwuU4lUHBuStPCimfbYE9quoZpLUG8A7Zb0bppqofFna6xhhjjMmfVWAYY4wx5qIQkRk43VlO4LRuqE52V5XNQLSnW477atlTOC1DfgYqkN2NIwOnVceKC5TPJTitRY7jtP6oTnar1bdV1W+rF2OMMcZcWNaFxBhjjDEXywb3k4FTKZCC06VjLD6VF67TwDycbh5VcLqi/IjTbabFhaq8cK1y8yU4bxlJcvM9HLjjAqZrjDHGmHxYCwxjjDHGGGOMMcYUe9YCwxhjjDHGGGOMMcWeVWAYY4wxxhhjjDGm2LMKDGOMMcYYY4wxxhR7VoFhjDHGGGOMMcaYYs8qMIwxxhhjjDHGGFPsWQWGMcYYY4wxxhhjij2rwDDmHIjIXBGZWtT5uNBEJFZERhRBum1FZI+IJItIn4ud/qVCRGaJSEwhx3m3iGwuzDiNMcaYS4GItBeRH4o6H8aYs7MKDFNkRCRORG4q6nwUFhG5TET+JSKH3B/g+0XkGZ/l57S951NZIiITRWSjn/nVROSMiIT+lngvgieBmapaXlWXi4iKSL3CilxErhKRBSJyRESSROQTEWmVK8ydInJARH4VkeUiUsVn2eUi8pqInBSRn0RkbK51I0TkSxFJcf9GFFbefanqaFX924WI2xhjjLmYikN5UFU3qWrDCxW/iHQVkY0ickpEjorIBhHpdaHSM+b3zCowzO+aiARdxOQmAs2BlkAFoBPw9UVM39d8oI2IXJtr/h3ADlX9bxHkqSDqAN8VRkQiUtLP7PLAF0AzoAowD1gpIuXddUKAl4DBQHUgBfg/n/UnA/XdfHYCxonIze66lwHvAm8Ald2433XnG2OMMaaIXOTyYO60bwcWA68DwTjliyeAW35DXCIi9vvN/KHZP4ApdkSkhIhMEJF9InJcRN7O9RR8sfv0O8mtzQ7xWTZXRP4jIqtE5Fegk1uz/6iIfOuus0hESvus01NEtovICRH5VETCfZY1EZGv3BrzRYB3PT9aAMtU9Yg64lT1dTee+cA1wHtu64xx+W2LiIwCBuH8QE4Wkffc+TVFZKlbe79fRP7iLyOqegj4GOeHuK8hwDwRqSwi77vxJLrfgwMcj8ki8obPdF23ZURJd7qiiMwWkXgROSwiUwMVFESkpYh85u7reBGZ6fmBLyL7gOt89tFn7mrfuNMD3HD5Ha84ERkvIt8Cv+auxFDV/6nqv1U1XlUzVfVl4DLA89RlEPCeqm5U1WQgBugrIhV89t/fVDVRVb8HXgHudpdFASWBZ1X1tKo+DwhwY4B9cbmIzBCRH0XkZ3G6hZRxl0WJ05LncRE55m7XIJ91va1zxGlV8767PxJEZJOncCMiN4jTHeiEiHzn+7RHRKqKyApxWpNsBf6UK3/Xi8haN84fRKS/z7LuIrLT/b84LCKP+ttGY4wx5re6mOVBz33XZ/2zlR3HueWYIyIyQgK0GBURAf6NU3Z4VVWTVDVLVTeo6kg3zNnKWbEi8ncR+QTnwcrjIrItVzoPi8gK93vA8oUxvwdWgWGKo78AfYCOQE0gEXjRZ/kHOE/BrwK+At7Mtf6dwN9xWkF4+vT3B24GrgXCcX90ikhT4DXgXqAqztP3Fe7F/zJgOU5rhio4tee35ZPvz4GxInK/iIS5Ny0AVHUw8CNwi9s94p/5bYv7w/pN4J9u+FvcH6XvAd8AtYBo4CER6RogP/PwqcAQkYZABLAA539/Dk5LgmuAVGBmPtuWn3lABlAPaAJ0AQKNn5EJPAxUA1q723A/gKr+iZz7qLW7TmN3elF+x8snjYFAD6CSqmbkl3FxunhcBux1Z4Xg7F/cPO0DzgANRKQyzvn4jU8U37jreNb9VlXVZ/m3Pstz+wfQAOeY1MM5pk/4LL8aZz/VAoYCL7vHMLdHgEPAlThPdR4HVERK4Zwva3DOrz8Db/rE8SKQBtQAhrkfz34pB6wF3nLXHQj8n0/hcDZwr6pWAEJxKsuMMcaYwnTRyoMBBCo73gyMBW7CuX93zCeOhkBtYEk+YQpiMDAKZ1teABqKSH2f5Xfi3LPh7OULYy5pVoFhiqN7gb+q6iFVPY3TbP92T020qr6mqqd8ljUWkYo+67+rqp+4Ndxp7rzn3ZYRCTg/6jxjE4wEXlLVLe4T+XnAaSDS/ZTCeaKerqpLcLofBPI0zk1jELANOCwiQ/Pb0AJsi68WwJWq+qSqnlHV/+G0ALgjQPhlQHURaeNODwE+UNWjqnpcVZeqaoqqnsK5wed3A/ZLRKoD3YCHVPVXVf0FeCZQnlT1S1X9XFUzVDUOpwLiXNLN73h5PK+qB1U19Sx5vwKncmqKqia5s8sDSbmCJuEUGMr7TOdedrZ1c6ct7rY8rKoJ7jF4irz7LcZtzbEBWIlTmMotHacSoo57nm5yK1Ei3TxNc8+Xj4H3gYHitJC5DXjCPW7/xamI8ugJxKnqHPdYfQUsBW73SbORiFzhtkb5yk++jDHGmPNxMcuD/gQK2x+Yo6rfqWoKMCWfOKq6f+MLvNX+zXXTy3DLLO/iPFzArci4HueBTkHLF8ZcsqwCwxRHdYBlbrP3E8D3OE/uq4tIkIhMc5sTngTi3HWq+ax/0E+cP/l8TyH7x2gd4BFPWm56tXFq+msCh3M9UT8QKNPuD+oXVbUtUAmnUuA1EbnBX/gCbouvOkDNXHl9HOepu7/8pOC0Ghni3tAG4f5IFZGyIvKSOINVngQ2ApXk3PuI1sGp5In3ydNLOE9D/G1zA3G6O/zkpvtUPtsbKL1Ax8vD3/HPnY8yOIWRz1X1aZ9FycAVuYJfAZxyl5FruWfZ2dbN7UqgLPClz3Z86M73SFTVX32mD5BzOz2m47QgWSMi/xORCe78msBBVc3KFUctN52S5NxXvud2HaBVrv08CKdVCDiVH92BA+IMRNYaY4wxpnBdzPKgP4HC1swVd37ljuPu3xr5hCmI3Gm8hVuBgdP6Yrlb7itI+cKYS5pVYJji6CDQTVUr+XxKq+phnIt0b5xmexWBuu464rO+UnAHgb/nSqusqi7AqS2v5dsVBKe7xVmpaqqqvojT3LFRgHydbVtyhz8I7M+V1wqq2j2frMzDeVLQGaclwPvu/EdwmjW2UtUrgA650vb1K87N0ONqn+8HcVpAVPPJ0xWqGqjbxH+AXUB9N93HA6QZSH7HyyPf4+92N1kOHMZ5uuPrO6CxT9jrgMuB3aqaiHNONPYJ35jsQUe/A8JznS/h+B+U9BhOt50Qn+2oqKq+BanKblcOj2uAI7kjcp8+PaKq1+EMCDZWRKLdsLUl52Bf17jbfRSn20/tXMs8DgIbcu3n8qp6n5vmF6raG6eiajnwtp9tNMYYY87HxSwPnot4nME4PWoHCgj8gLMd+XVBzq+c5ZF7W9YA1dyusAPJ7j5SkPKFMZc0q8AwRa2UiJT2+ZQEZgF/F5E6ACJypYj0dsNXwPnBfBznYv/Ueab/CjBaRFqJo5yI9BBn0MbPcH7k/UVESopIX5w3jPglIg+JMwhUGTf8UDe/njeR/IwzSKXH2bYld/itwElxBqks4z59CBWRFvls3ybgBPAysFBVz/iknQqcEGdArEn5xLEd6CAi17hNMyd6FqhqPM5N9F8icoU4A279SUQCdQupAJwEkkXkeuC+fNKFvPsgv+N1VuKMC7EEZ9uH5GqdAE7/2VvEeR98OZzXur7jNsEEZwTx/yfOIKjX4zTTnOsui8V5MvQXccZQecCdn2d8CDfdV4BnROQqN2+1JO94JlPEeT1ve5xuHYv9bFNPEannVpycdPOQCWzBKRSNE5FSIhKFU8GxUFUzgXeAyW5rnEY442x4vI8z7sdgd91SItJCnEFBLxORQSJSUVXTfdI0xhhjfquiLg+ei7eBe9x7YlnyGV/CbcU7FogRkXt8ykrtRORlN1jAclY+8WbglGem44zTttadX9DyhTGXLKvAMEVtFc6PSc9nMvAcsAKnSfwpnMExW7nhX8dp6n4Y2Oku+81UdRvOj9CZOK0l9uIO0uT+2O/rTicCA3B+9AWSCvwLp8nhMWAMcJs6Y1WAM0bG/3Ob9D1agG2ZjTPOwAkRWe7+6LwFpw/mfjeNV3GePATaPnXTqeP+9XgWKOPG8TlO88JAcawFFuEMSPkl2a04PIbgDIS5E2c/LSFwU8lHcZ6anMK5wS4KlK5rMs5bU06ISP/8jlcBtcGpCOiCU3mT7H7aA6jqd8BonIqMX3AKSPf7rD8J2Idz3DYA01X1Q3fdMziDjQ3BqTQaBvTxqTTKbbyb/8/d5q/ryH4bCjjnUSJOS4o3gdGqustPPPXddZNxKt3+T1Vj3XR74YxRcgzndbBDfOJ4AKc57E84lTBzPBG6FTZdcPrMHnHD/AOnNQo4g4nFufkeDdwVYBuNMcaYgijS8uC5UNUPgOeB9Tj3cc9b004HCL8Epww5DOee+jMwFWcci4KUswJ5C6cFymLNOWj52coXxlzSJGf3fmOMMUXNbS3xhqr6fbWtMcYYY4oHccY6+y9wuZ7l7WfGmPNnLTCMMcYYY4wxpoBE5Fa3S2VlnBaK71nlhTEXxyVRgSEir4nILyLy3wDLRUSeF5G9IvKtiDS92Hk0xhhjjDkXInKziPzgll8m+FkeJSJJIrLd/QTsa2+MuajuxRkQex/OOFBnG9PLGFNILokuJCLSAad/9+uqGupneXfgzziv9WsFPKeqrXKHM8YYY4wpDsR5bfVunLdEHQK+AAaq6k6fMFHAo6ras0gyaYwxxhQzl0QLDFXdCCTkE6Q3TuWGqurnQCUROd/3LRtjjDHGXCgtgb2q+j930N2FOOUZY4wxxgRQsqgzUEhq4bxj2eOQOy8+d0ARGQWMAihXrlyz66+//qJk0BhjjLkQvvzyy2OqemVR58OcM39lF3+tR1uLyDc4by941H1bUh5WvjHGGPN7Eqh883upwBA/8/z2jVHVl4GXAZo3b67btm27kPkyxhhjLigROVDUeTC/SUHKLl8BdVQ12e0uuxzn1cl5V7TyjTHGmN+RQOWbS6ILSQEcAmr7TAfjPKkwxhhjjCmOzlp2UdWTqprsfl8FlBKRahcvi8YYY0zx8nupwFgBDHHfRhIJJKlqnu4jxhhjjDHFxBdAfRG5VkQuA+7AKc94icjVIiLu95Y45bbjFz2nxhhjTDFxSXQhEZEFQBRQTUQOAZOAUgCqOgtYhfMGkr1ACnBP0eTUGGOMMebsVDVDRB4AVgNBwGuq+p2IjHaXzwJuB+4TkQwgFbhDL4XXxxljjDEXyCVRgaGqA8+yXIExhZFWeno6hw4dIi0trTCiM8aY373SpUsTHBxMqVKlijorxlxS3G4hq3LNm+XzfSYw82Lny/x2Vo40xphzc67lyEuiAuNiOnToEBUqVKBu3bq4rTaNMcYEoKocP36cQ4cOce211xZ1dowxpkhZOdIYYwrut5Qjfy9jYBSatLQ0qlatajcdY4wpABGhatWq9rTRGGOwcqQxxpyL31KOtAoMP+ymY4wxBWfXTGOMyWbXRGOMKbhzvWZaBYYxxhhjjDHGGGOKPavAMAF1796dEydOnHc827dvZ9WqVWcPeB7i4uIIDQ29oGmczVNPPVWo4S41sbGxfPrpp97pWbNm8frrrxd6On/961+pXbs25cuXzzH/9OnTDBgwgHr16tGqVSvi4uK8y+bNm0f9+vWpX78+8+bN+03pXozz+MiRI9x+++2FGmdCQgKdO3emfv36dO7cmcTERL/hPvzwQxo2bEi9evWYNm2ad/7ixYsJCQmhRIkSbNu2rVDzZowx5vfLypHnxsqRVo48X3+UcqRVYPxBZGRknPM6q1atolKlSued9sX4hy0OLsUbT2ZmZqHFlfvGM3r0aIYMGVJo8XvccsstbN26Nc/82bNnU7lyZfbu3cvDDz/M+PHjAefCO2XKFLZs2cLWrVuZMmVKwItvfi7GeVyzZk2WLFlSqHFOmzaN6Oho9uzZQ3R0dI6bikdmZiZjxozhgw8+YOfOnSxYsICdO3cCEBoayjvvvEOHDh0KNV/GGGMuHVaOvPCsHGnlyPP1hylHquof9tOsWTPNbefOnXnmXWy9e/fWpk2baqNGjfSll17yzi9XrpyOHTtWmzRpojfeeKP+8ssvqqrasWNHffDBB7V169YaEhKiW7ZsUVXVSZMm6ciRI7Vz5846cOBAjYuL0xtvvFHDwsL0xhtv1AMHDuiJEye0QYMGumvXLlVVveOOO/Tll19WVdU6dero0aNHdf/+/dqwYUMdPny4hoSE6J133qlr167VNm3aaL169bzpbdmyRVu3bq0RERHaunVr3bVrl54+fVpr166t1apV08aNG+vChQs1OTlZ77nnHm3evLlGRETo8uXL8+yD/v3768qVK73TQ4cO1SVLluj+/fu1Xbt22qRJE23SpIl+8sknqqq6f/9+DQkJUVXVOXPm6JgxY7zr9ujRQ9evX6+qqqtXr9bIyEht0qSJ3n777Xrq1ClVVR0/frzecMMNGhYWpo888ki+x+fIkSPavn17bdy4sYaEhOjGjRt1/PjxWqJECW3cuLHeeeedAY+jv3Dz58/XFi1aaOPGjXXUqFGakZGRJ80pU6Zo8+bNNSQkREeOHKlZWVmqqrp161YNCwvTyMhIffTRR7374Ndff9V+/fppWFiY9u/fX1u2bKlffPGF9zyKiYnRli1b6qZNmwKm/+qrr2r9+vW1Y8eOOmLECO8+XbFihbZs2VIjIiI0Ojpaf/rpJ92/f79Wr15da9asqY0bN9aNGzfqpEmTdPr06aqq+vXXX2urVq00LCxM+/TpowkJCd5zd9y4cdqiRQutX7++bty4Md9976tcuXI5prt06aKffvqpqqqmp6dr1apVNSsrS9966y0dNWqUN9yoUaP0rbfeyjfut99+W0NCQjQ8PFzbt29/TufxnDlztFevXtq1a1dt0KCBTp482W8asbGx2rhxY23cuLFGREToyZMnc5zHw4cP9y6vVq2aN55//vOf2rx5cw0LC9MnnnjirPupQYMGeuTIEVV1zt0GDRrkCfPpp59qly5dvNNPPfWUPvXUUznCdOzY0XsO+VMcrp1/VMA2LQb3VPsUn4+/8o25OIrDtdDKkVaOzM3KkXlZObL4lyMDlW+K/CZblJ+zVWBMXvFf7T/r00L9TF7x34AHzuP48Z2WYjAAACAASURBVOOqqpqSkqIhISF67Ngxz0HUN954Q1WdC5HnQuC5MKiqbtiwwXviTpo0SZs2baopKSmqqtqzZ0+dO3euqqrOnj1be/furaqqa9as0cjISF2wYIF27drVmw/fG09QUJB+++23mpmZqU2bNtV77rlHs7KydPny5d54kpKSND09XVVV165dq3379lXVvDeCiRMn6vz581VVNTExUevXr6/Jyck59sE777yjQ4YMUVXV06dPa3BwsKakpOivv/6qqampqqq6e/du9RzDgtx4jh49qu3bt/emNW3aNJ0yZYoeP35cGzRo4L2YJyYmqqrqu+++qzExMXmOz4wZM3Tq1KmqqpqRkaEnT55U1bwXwkDH0Tfczp07tWfPnnrmzBlVVb3vvvt03rx5edL0xKWqetddd+mKFStUVTUkJMR78x0/frx3H0yfPt17sd2xY4cGBQV5LxqALlq0KN/0Dx8+rHXq1NHjx4/rmTNntF27dt59mpCQ4N1Xr7zyio4dO1ZVNceNJvd0WFiYxsbGqqpqTEyMPvjgg6rqnLue9VeuXKnR0dGqqnr48GHt1q1bnv3gK/f+DgkJ0YMHD3qnr7vuOj169KhOnz5d//a3v3nnP/nkkzny6U9oaKgeOnRIVbPPh4Kex3PmzNGrr75ajx075j32/i7YPXv21M2bN6uq6qlTpzQ9PT3HeewRFxenDRs21Li4OF29erW34JGZmak9evTQDRs2qKpqt27d9PDhw3nSqVixYo7pSpUq5QmzePFiHT58uHf69ddfz7GtqlaBUZxZBYZ9cn+sAqPoWDnSypFWjrRypJUjC6ccGah8U7Lw2nKYwvL888+zbNkyAA4ePMiePXuoWrUqJUqUYMCAAQDcdddd9O3b17vOwIEDAejQoQMnT5709jns1asXZcqUAeCzzz7jnXfeAWDw4MGMGzcOgM6dO7N48WLGjBnDN9984zdP1157LWFhYQCEhIQQHR2NiBAWFubtI5aUlMTQoUPZs2cPIkJ6errfuNasWcOKFSuYMWMG4Lxy7Mcff+SGG27whunWrRt/+ctfOH36NB9++CEdOnSgTJkyJCUl8cADD7B9+3aCgoLYvXt3gffr559/zs6dO2nbti0AZ86coXXr1lxxxRWULl2aESNG0KNHD3r27Ondd7169coTT4sWLRg2bBjp6en06dOHiIgIv+kFOo6+PvroI7788ktatGgBQGpqKldddVWeuNavX88///lPUlJSSEhIICQkhPbt23Pq1CnatGkDwJ133sn7778PwObNm3nwwQcBp+lWeHi4N66goCBuu+22fNPfunUrHTt2pEqVKgD069fPu68PHTrEgAEDiI+P58yZM2d9Z3NSUhInTpygY8eOAAwdOpR+/fp5l3vO42bNmnnPpZo1a55zMzvnOpeTiAScn5+2bdty9913079//xz/Z74Cncfg/E95jnXfvn3ZvHkzzZs3z5PG2LFjGTRoEH379iU4ODhPGmlpafTr14+ZM2dSp04dXnjhBdasWUOTJk0ASE5OZs+ePXTo0OG8miX+ln1kjDGmeLJypJUjc7Ny5NlZOfLSKUdaBUY+Jt0SctHTjI2NZd26dXz22WeULVuWqKiogO/F9T0xcp8knuly5coFTMsTJisri++//54yZcqQkJDg9x/g8ssv934vUaKEd7pEiRLefpExMTF06tSJZcuWERcXR1RUlN90VZWlS5fSsGHDgHkrXbo0UVFRrF69mkWLFnlvrM888wzVq1fnm2++ISsri9KlS+dZt2TJkmRlZXmnPftPVencuTMLFizIs87WrVv56KOPWLhwITNnzuTjjz8OmLcOHTqwceNGVq5cyeDBg3nsscfy9NEr6HFUVYYOHcrTTz8dML20tDTuv/9+tm3bRu3atZk8eTJpaWl+Lxa+8QZSunRpgoKC8k3fc8P0589//jNjx46lV69exMbGMnny5IBhC8JzLgUFBf2mPrYewcHBHDx4kODgYDIyMkhKSqJKlSoEBwcTGxvrDXfo0KGA56bHrFmz2LJlCytXriQiIoLt27fnCRPoPN6yZYvf/8cXX3yRV155BXD6BU+YMIEePXqwatUqIiMjWbduXZ7zefTo0fTt25ebbrrJm+bEiRO59957C7pbqF69OvHx8dSoUYP4+Hi/BRvPvvM4dOgQNWvWLHAaJqcszeLbo9+SkpECQEkpSZPqTShVolQR58wYczFZOTKblSOzWTnSypFg5cjzYYN4FjNJSUlUrlyZsmXLsmvXLj7//HPvsqysLO/ALG+99Rbt2rXzLlu0aBHg1JhWrFiRihUr5om7TZs2LFy4EIA333zTu/4zzzzDDTfcwIIFC7w1wr8177Vq1QJg7ty53vkVKlTg1KlT3umuXbvywgsveC+OX3/9td/47rjjDubMmcOmTZvo2rWrN40aNWpQokQJ5s+f73fwoLp167J9+3aysrI4ePCgd6CeyMhIPvnkE/bu3QtASkoKu3fvJjk5maSkJLp3786zzz7r9yLj68CBA1x11VWMHDmS4cOH89VXXwFQqlQp777L7zj6houOjmbJkiX88ssvgDNQ0IEDB3Kk57lhVatWjeTkZO85ULlyZSpUqOCN23NsAdq1a8fbb78NwM6dO9mxY4ffbQmUfsuWLdmwYQOJiYlkZGSwdOlS7zq+x9l3JObcx9mjYsWKVK5cmU2bNgEwf/58by16YerVq5c3P0uWLOHGG29EROjatStr1qwhMTGRxMRE1qxZ4z2fJk6c6Pcmu2/fPlq1asWTTz5JtWrVOHjw4Dmdx2vXriUhIYHU1FSWL19O27ZtGTNmDNu3b2f79u3UrFmTffv2ERYWxvjx42nevDm7du3KkYcXX3yRU6dOMWHChBxpvvbaayQnJwNw+PBh77EryH6ZN28evXv3zhOmRYsW7Nmzh/3793PmzBkWLlzo96mRKZgvf/6SwR8M5t6193Lv2nsZvmY48777baOWG2PMubByZDYrRzqsHFkwVo48+34pLuVIq8AoZm6++WYyMjIIDw8nJiaGyMhI77Jy5crx3Xff0axZMz7++GOeeOIJ77LKlSvTpk0bRo8ezezZs/3G/fzzzzNnzhzCw8OZP38+zz33HLt37+bVV1/lX//6F+3bt6dDhw5MnTr1N+V93LhxTJw4kbZt2+a4IXTq1ImdO3cSERHBokWLiImJIT09nfDwcEJDQ4mJifEbX5cuXdi4cSM33XQTl112GQD3338/8+bNIzIykt27d/t9MtC2bVtvU8VHH32Upk2bAnDllVcyd+5cBg4cSHh4OJGRkezatYtTp07Rs2dPwsPD6dixI8888wwAK1asyLGPPWJjY4mIiKBJkyYsXbrU28Ru1KhRhIeHM2jQoHyPo2+4Ro0aMXXqVLp06UJ4eDidO3cmPj4+R3qVKlVi5MiRhIWF0adPH28zPXBGTB41ahStW7dGVb0Fjvvvv5+jR48SHh7OP/7xD8LDw/0WRgKlX6tWLR5//HFatWrFTTfdRKNGjbzrT548mX79+tG+fXuqVavmjeuWW25h2bJlREREeG8yHvPmzeOxxx4jPDyc7du3+92vvo4cOUL37t39Lhs3bhzBwcGkpKQQHBzsrbkfPnw4x48fp169evz73//2jpJcpUoVYmJiaNGiBS1atOCJJ57wNmncsWMHV199dZ40HnvsMcLCwggNDaVDhw40btz4nM7jdu3aMXjwYCIiIrjtttvyNPsDePbZZwkNDaVx48aUKVOGbt265Vg+Y8YMduzYQUREBBEREcyaNYsuXbpw55130rp1a8LCwrj99tu9N8Pu3btz5MiRPOlMmDCBtWvXUr9+fdauXeu9kfnu45IlSzJz5ky6du3KDTfcQP/+/QkJcZ4cLlu2jODgYD777DN69OjhvWmbwBLSEgB4qt1TvN7tdVrVaMWc/84h5pMY7+ebo/6bWRtjzPmwcmQ2K0c6rByZk5UjL/1ypOTXROj3rnnz5pr7fbTff/99jj50xUn58uW9NWa+oqKimDFjht+T2/y+JScne99jPW3aNOLj43nuuefIzMwkPT2d0qVLs2/fPqKjo9m9e7f3Bn4ucWdkZHDrrbcybNgwbr311gu1KUWia9eurF69ulDjnDt3Ltu2bWPmzJmFGm9xV5yvnRfb8r3Lifkkhg/6fkBwhWC+O/4d4zaM40zWGW+YCS0nEH1NdKGkJyJfqqrdAIyXv/KNuTiK87XQypEmNytHnh8rRxYef9fOQOUbGwPDmEvYypUrefrpp8nIyKBOnTreJpcpKSl06tSJ9PR0VJX//Oc/53TTAaeGfN26daSlpdGlSxf69OlzAbagaBX2TccYgNSMVADKlioLQEjVEFb2XVmUWTLGGGPysHLk+bFyZNGwFhiXUAsMY4wpruzamW32jtk8+9WzbB20lTIly+RYlpWaStJ771Ghc2dKVq5cKOlZCwyTm7XAKDp2LTTGmHNnLTCMMcaYIpKakYoglA7KHg08/aefSHzzLU68/TaZSUmgUHlA/yLMpTHGGGPMpccqMIwxxphClJqR6m15kfLV1yS8+n+cWr8ZgArXXUaVzldQJiLvYGjGGGOMMSZ/VoFhjDHGFKK0tGQ6fgdx/fqT9t//UqJMKao0TKFy1+ZcVsl5Xz2Xly/aTBpjjDHGXIKsAsMYY4wpBJnJv3JiyWK6vbSCcolpZF19mOqDo6gUtJ4S10TCnYuKOovGGGOMMZe0EkWdAVN8de/enRMnTpx3PNu3b2fVqlWFkKPA4uLiCA0NvaBpnM1TTz1VqOEuNbGxsXz66afe6VmzZvH6668Xejp//etfqV27tve1Xx6nT59mwIAB1KtXj1atWhEXF+ddNm/ePOrXr0/9+vWZN2/eb0r3YpzHR44c4fbbby/UOBMSEujcuTP169enc+fOJCYm+g334Ycf0rBhQ+rVq+d99zk4o4jXqlXL+x7xC70PLkUZx47xy7PPsvfGG/ll2j9IqhzE67dmcV3H76iS/hYl0uKh8cCizqYxxlxUVo48N1aOtHLk+fqjlCOtAuMPIiMj45zXWbVqFZUqVTrvtC/GP2xxcCneeDIzMwstrtw3ntGjRzNkyJBCi9/jlltuYevWrXnmz549m8qVK7N3714efvhhxo8fDzgX3ilTprBlyxa2bt3KlClTAl5883MxzuOaNWuyZMmSQo1z2rRpREdHs2fPHqKjo3PcVDwyMzMZM2YMH3zwATt37mTBggXs3LnTu/zhhx9m+/btbN++ne7duxdq/i5lZw4cIH7yZPbeGM3xl16mXKtW1F20kEXDa3CwriKjN8G4/TDhIIT8/l4fZ4z547By5IVn5UgrR56vP0o50iowiqE+ffrQrFkzQkJCePnll73zy5cvzyOPPELTpk2Jjo7m6NGjAERFRfHQQw/Rpk0bQkNDvf+UkydPZtSoUXTp0oUhQ4Zw4MABoqOjCQ8PJzo6mh9//JGkpCQaNmzIDz/8AMDAgQN55ZVXAKhbty7Hjh0jLi6O66+/nhEjRhAaGsqgQYNYt24dbdu2pX79+t70tm7dSps2bWjSpAlt2rThhx9+4MyZMzzxxBMsWrSIiIgIFi1axK+//sqwYcNo0aIFTZo04d13382zDwYMGJDjn/zuu+9m6dKlxMXF0b59e5o2bUrTpk1zXOg85s6dywMPPOCd7tmzJ7GxsQCsWbOG1q1b07RpU/r160dycjIAEyZMoFGjRoSHh/Poo4/me3zi4+Pp0KEDERERhIaGsmnTJiZMmEBqaioREREMGjQo4HH0F+6NN96gZcuWREREcO+99/q9GTz55JO0aNGC0NBQRo0ahef1x1988QXh4eG0bt2axx57zPv0ICUlhf79+xMeHs6AAQNo1aoVnlfqlS9fnieeeIJWrVrx2WefBUx/9uzZNGjQgKioKEaOHOndp++99x6tWrWiSZMm3HTTTfz888/ExcUxa9YsnnnmGSIiIti0aROTJ09mxowZgHPRjoyMJDw8nFtvvdV74Y+KimL8+PG0bNmSBg0asGnTpnz3PUBkZCQ1atTIM//dd99l6NChANx+++189NFHqCqrV6+mc+fOVKlShcqVK9O5c2c+/PDDfNNYvHgxoaGhNG7cmA4dOpzTeTx37lx69+7NzTffTMOGDZkyZYrfNDZs2OCtjW7SpAmnTp3K8QRoxIgR3uVXXnmlN57p06fTokULwsPDmTRp0ln3l+9+GTp0KMuXL88TZuvWrdSrV4/rrruOyy67jDvuuMPv/6VxpO3axaGHHmZft+4kLX2Hir17c93KlQS/8DxlGjcmJSOFMpoFVa6FslWg9BVFnWVjzB+IlSOtHJmblSOzWTnyd1COVNU/7KdZs2aa286dO7MnVo1Xfa174X5Wjc+TZm7Hjx9XVdWUlBQNCQnRY8eOqTpXGn3jjTdUVXXKlCk6ZswYVVXt2LGjjhgxQlVVN2zYoCEhIaqqOmnSJG3atKmmpKSoqmrPnj117ty5qqo6e/Zs7d27t6qqrlmzRiMjI3XBggXatWtXbz7q1KmjR48e1f3792tQUJB+++23mpmZqU2bNtV77rlHs7KydPny5d54kpKSND09XVVV165dq3379lVV1Tlz5njzqqo6ceJEnT9/vqqqJiYmav369TU5OTnHPnjnnXd0yJAhqqp6+vRpDQ4O1pSUFP311181NTVVVVV3796tnmO4f/9+73bnTq9Hjx66fv16PXr0qLZv396b1rRp03TKlCl6/PhxbdCggWZlZXnzpKr67rvvakxMTJ7jM2PGDJ06daqqqmZkZOjJkydVVbVcuXIFOo6+4Xbu3Kk9e/bUM2fOqKrqfffdp/PmzcuTpicuVdW77rpLV6xYoaqqISEh+sknn6iq6vjx4737YPr06Tpq1ChVVd2xY4cGBQXpF198oarOebRo0aJ80z98+LDWqVNHjx8/rmfOnNF27dp592lCQoJ3X73yyis6duxYVXXOt+nTp3vz6TsdFhamsbGxqqoaExOjDz74oKo6565n/ZUrV2p0dLSqqh4+fFi7deuWZz/4yr2/Q0JC9ODBg97p6667To8eParTp0/Xv/3tb975Tz75ZI58+hMaGqqHDh1S1ezzoaDn8Zw5c/Tqq6/WY8eOeY+9Z9/76tmzp27evFlVVU+dOqXp6ek5zmOPuLg4bdiwocbFxenq1at15MiRmpWVpZmZmdqjRw/dsGGDqqp269ZNDx8+nCedihUr5piuVKlSnjCLFy/W4cOHe6dff/1177ZOmjRJ69Spo2FhYXrPPfdoQkKC332W49r5O5OZlamLdi3St5ZP1U139dCdDa/X75s1059n/EvP/PxznvB932irf/6/eqru/8mFBmzTYnBPtU/x+fgr35iLw8qRVo60cqSVI60cWTjlyEDlGxvEsxh6/vnnWbZsGQAHDx5kz549VK1alRIlSjBgwAAA7rrrLvr27etdZ+BAp391hw4dOHnypLfPYa9evShTxnmd32effcY777wDwODBgxk3bhwAnTt3ZvHixYwZM4ZvvvnGb56uvfZawsLCAAgJCSE6OhoRISwszNtHLCkpiaFDh7Jnzx5EhPT0dL9xrVmzhhUrVnhrVdPS0vjxxx+54YYbvGG6devGX/7yF06fPs2HH35Ihw4dKFOmDElJSTzwwANs376doKAgdu/eXeD9+vnnn7Nz507atm0LwJkzZ2jdujVXXHEFpUuXZsSIEfTo0YOePXt6912vXr3yxNOiRQuGDRtGeno6ffr0ISIiwm96gY6jr48++ogvv/ySFi1aAJCamspVV12VJ67169fzz3/+k5SUFBISEggJCaF9+/acOnWKNm3aAHDnnXfy/vvvA7B582YefPBBAEJDQwkPD/fGFRQUxG233ZZv+lu3bqVjx45UqVIFgH79+nn39aFDhxgwYADx8fGcOXOGa6+9Nt/9npSUxIkTJ+jYsSPg1N7269fPu9xzHjdr1sx7LtWsWfOcm9k517mcRCTg/Py0bduWu+++m/79++f4P/MV6DwG53/Kc6z79u3L5s2bad68eZ40xo4dy6BBg+jbty/BwcF50khLS6Nfv37MnDmTOnXq8MILL7BmzRqaNGkCQHJyMnv27KFDhw7n1Swxv3103333ERMTg4gQExPDI488wmuvvfab07oU7f5kFWemTabZXiW5NLzdrgSVB9/GXzqO9Rs+Nes0ZUuUgrOcZ8YYcyFYOdLKkblZOfLsrBx56ZQjrQIjP93y9vG50GJjY1m3bh2fffYZZcuWJSoqirS0NL9hff95cv8jeabLlSsXMC1PmKysLL7//nvKlClDQkKC33+Ayy+/3Pu9RIkS3ukSJUp4+0XGxMTQqVMnli1bRlxcHFFRUX7TVVWWLl1Kw4YNA+atdOnSREVFsXr1ahYtWuS9sT7zzDNUr16db775hqysLEqXLp1n3ZIlS5KVleWd9uw/VaVz584sWLAgzzpbt27lo48+YuHChcycOZOPP/44YN46dOjAxo0bWblyJYMHD+axxx7L00evoMdRVRk6dChPP/10wPTS0tK4//772bZtG7Vr12by5MmkpaX5vVj4xhtI6dKlCQoKyjd9zw3Tnz//+c+MHTuWXr16ERsby+TJkwOGLQjPuRQUFPSb+th6BAcHc/DgQYKDg8nIyCApKYkqVaoQHBzsbfoJzo0z0LnpMWvWLLZs2cLKlSuJiIhg+/btecIEOo+3bNni9//xxRdf9DarXbVqFRMmTKBHjx6sWrWKyMhI1q1bl+d8Hj16NH379uWmm27ypjlx4kTuvffegu4WqlevTnx8PDVq1CA+Pt5vwcaz7zwOHTpEzZo1vet7jBw50lswK26yNItMzW42W1JKnrWAcTap27dz9P/+D924iYalIW1YX0JHPczTG0dRI+NQ4PWyMigTlPfaZIz5g7FypJeVI7NZOdLKkWDlyPNhY2AUM0lJSVSuXJmyZcuya9cuPv/8c++yrKws78Asb731Fu3atfMuW7TIeT3f5s2bqVixIhUrVswTd5s2bVi4cCEAb775pnf9Z555hhtuuIEFCxZ4a4R/a95r1aoFOP23PCpUqMCpU6e80127duWFF17wXhy//vprv/HdcccdzJkzh02bNtG1a1dvGjVq1KBEiRLMnz/fbz+/unXrsn37drKysjh48KC3b2VkZCSffPIJe/fuBZz+fbt37yY5OZmkpCS6d+/Os88+6/ci4+vAgQNcddVVjBw5kuHDh/PVV18BUKpUKe++y+84+oaLjo5myZIl/PLLL4AzUNCBAwdypOe5YVWrVo3k5GTvOVC5cmUqVKjgjdtzbAHatWvH22+/DcDOnTvZsWOH320JlH7Lli3ZsGEDiYmJZGRksHTpUu86vsfZdyTm3MfZo2LFilSuXNnbL3H+/PneWvTC1KtXL29+lixZwo033oiI0LVrV9asWUNiYiKJiYmsWbPGez5NnDjR70123759tGrViieffJJq1apx8ODBczqP165dS0JCAqmpqSxfvpy2bdsyZswY7wBGNWvWZN++fYSFhTF+/HiaN2/Orl27cuThxRdf5NSpU0yYMCFHmq+99pq3z+3hw4e9x64g+2XevHn07t07T5gWLVqwZ88e9u/fz5kzZ1i4cKH3qVF8fLw33LJly4p8lPZA+r/Xn6bzm3o/f9/y9xzLdxzdwdoDa1l7YC3rDqzjowMf8dGPzufjHz9m/Y/rWf/jemIPxvLpunn8b9jdxN0xkLRvd5AwtBtj7g+i0r0jKFOpGn+q+Cf2ndgXMC8pmkmZklaBYYy5+Kwcmc3KkQ4rRxaMlSPPvl+KSznSWmAUMzfffDOzZs0iPDychg0bEhkZ6V1Wrlw5vvvuO5o1a0bFihW9NxtwLkJt2rTh5MmTAZvlPP/88wwbNozp06dz5ZVXMmfOHHbv3s2rr77K1q1bqVChAh06dGDq1KkBB4zJz7hx4xg6dCj//ve/ufHGG73zO3XqxLRp04iIiGDixInExMTw0EMPER4ejqpSt25db5M1X55Bo3r16sVll10GwP33389tt93G4sWL6dSpk98nA23btvU2VQwNDaVp06YAXHnllcydO5eBAwdy+vRpAKZOnUqFChXo3bu3tzb6mWeeAWDFihVs27aNJ598Mkf8sbGxTJ8+nVKlSlG+fHnvK55GjRpFeHg4TZs25bXXXgt4HH3Dvfnmm0ydOpUuXbqQlZVFqVKlePHFF6lTp443fKVKlRg5ciRhYWHUrVvX20wPnAGSRo4cSbly5YiKivIWOO6//36GDh1KeHg4TZo0ITw83G9hpFGjRn7Tj4yM5PHHH6dVq1bUrFmTRo0aedefPHky/fr1o1atWkRGRrJ//37AGdX59ttv59133+WFF17Ikc68efMYPXo0KSkpXHfddcyZMydPXnwdOXKEESNG+G3ONm7cON566y1SUlIIDg5mxIgRTJ48meHDhzN48GDq1atHlSpVvDfiKlWqEBMT491vTzzxhLdJ444dO/w273zsscfYs2cPqkp0dDSNGzfmmmuuKfB53K5dOwYPHszevXu588478zT7A3j22WdZv349QUFBNGrUiG7duuW4yM+YMYNSpUp5m5aOHj2a0aNH8/3339O6dWvAGUjrjTfe4KqrrqJ79+68+uqr3hpvjwkTJtC/f39mz57NNddcw+LFi/Ps45IlSzJz5ky6du1KZmYmw4YNIyQkxLu/t2/fjohQt25dXnrppXyPXVFIz0rnh8QfiKwRSYurW/Devvf47th33uVpGWkM+XAIGVn5P5kJPqoM2JhFq93KqfKXE/zoI1QeOJAv9rxB6jdrqTIzErKU6yqWZ3XFcqRmpFKmZJkccagqqShlSpW9INtqjDH5sXJkNitHOqwcmZOVIy/9cqTk10To96558+bqGVHX4/vvv8/Rh644KV++vLfGzFdUVBQzZszwe3KbAEg15AAAIABJREFU37fk5GTve6ynTZtGfHw8zz33HJmZmaSnp1O6dGn27dtHdHQ0u3fv9t7AzyXujIwMbr31VoYNG8att956oTalSHTt2pXVq1cXapxz585l27ZtzJw5s1DjLe6K8tp5LPUYnd7uxOOtHmfg9QOZuGkiX//yNR/e5owSfujUIbq9040HIh4gqnaUdz3FHRBq/16Y9DDsKwmllLUthANR1/DvIWsAeCF2Aq/Gvc/X5VpQotI1rPnvfB6pUo5udbtR4bIKiAiNr2xM1TJVycjKYMxHY3iozJ8Y3j/vSN0Xgoh8qap2AzBe/so35uKwcqS5lFg58vxYObLw+Lt2BirfWAsMYy5hK1eu5OmnnyYjI4M6dep4m1ympKTQqVMn0tPTUVX+85//nNNNB5wa8nXr1pGWlkaXLl3o06fPBdiColXYNx1z4aRmpHI89bh3+qqyV3FZkHNOn0hzBpurfHllAKqUrkJCWoI37PE0Z70bqt5AwyrZfU3Tf/mF47NeIvHtRYiWpHL0DVTt15X/7H6eeEnyhjuReoxKWVmUaPsg1IygyY+buUYT2PLTFgDOZJ5h0Q/ZTzIBqpauUpibb4wxxhQ6K0eeHytHFg2rwLiE+Ks1B3IMLGP+WAYMGOAdUdxXhQoVON+nb55Rkc25ufvuu7n77ruLOhu/O3d/eDc7j+/0Treu0ZqXu7wMQOJp533wlUpXApwKjNSMVFLSUyhbqqy34qNamWoAZJ46xfGXXyFh/nw0I4NKbRtSreLHlJr8KZStwtX/m8NXWdmDpSWmJVI5MxPKOiOCX1mlPiv3b4SxG5z44r9l17yunFGni0pJhUYN7EmmMaZ4sXKkyc3KkcWPlSPPziowjDHGFHvxyfG0uroVt/zpFt7b9x67E7NffZd02mkt4dsCA5yKjbKlynIs9ZgzP+VXEt6ZybFX55OZdJIrevbkyj8/wGVf/h32Xw1lnfVqXHYFJzNT+PX/s3ff4VFV6QPHv3cmvU96JQmQUFKBBAhV6U1AmtLUBWVdXfW3rIquC3Zl1V27YgUEKS4osoIKKIiAQKiioZPQkkBI75nJnN8fNwwtIGggIbyf58kD98659557Z3LvyTvnvMdciqu9K/lVRXhVW20BDLybws75YC4He2eMu5cSU1kBPaeBZgCDPSTcfq0ujRBCCCHEDUMCGEIIIRq8MksZrXxaMaT5EE6UnWBT9iYqLBU42TnZemB4OuoJwk4HMPLK8whxCyG39ARJ+6yUvDUGc7EdLv6V+A8x4hy1A1beDSd3Q3iK7VhBTt5Qmk12aTbNvJpRYCklQmngUJOY07up/u+2OeAeAGlLIawDdPnbtbsgQgghhBA3IAlgCCGEaNCqrdVUVlfiYqcHEIJcgwDIKs0i0jPSlgPDa/t8MNjhXVEznVtFHuW7fiHmyYXctN+K5udB6KSBuCVGoGWshXJ9O0KTIGmi7XiBLgFQmsbHuz4m1D2ULEsZbQxnjf0N0DNr8/UjZ9b1feEqnb0QQgghhDhNAhhCCCEatHJLOYBtytIQN33++MySTD2AUVmAs1XhtOopAEx2Rnw8g3F8fgYZa3bg5mrHlz2tTHnoZbTomqn5kidc9HhNPSPxyl7F0kNLAdCAGMNZ06L6t4KHfoaqmvHkmhF8o+ruhIUQQgghRK0M9V0B0XANGDCAgoKCP7yfHTt21DoPc13KyMggNjb2qh7jt7zwwuV9A3u55a43a9asYcOGDbblGTNm2OY2r0tPPPEEYWFhtmm/TqusrOS2226jefPmdOjQgYyMDNtrs2fPJioqiqioKGbPnv27jnstPseZmZmMGDGiTveZl5dH7969iYqKonfv3uTn59dabsKECfj7+9f771Ftzg9gBLvpc5RnlmYCUFB6Qk+y2etprH87gCG7Pa+/V43Lup383K85j9/nzP44K5pvxGUdz9MjjLVHjvNz///y8x0/s7PKj+FOIecWMoXrPTECYsC/JRiMdXOyQgjRSEg78spIO1LakX/UjdKOlADGDcJisVzxNsuXL8fLy+sPH/ta/MI2BNfjg6e6urrO9nX+g+fee+/ljjvuqLP9n3bLLbewefPmC9Z/9NFHmEwmDhw4wN/+9jemTJkC6Dfep59+mk2bNrF582aefvrpi958L+VafI6Dg4NZtGhRne5z+vTp9OzZk/3799OzZ0+mT59ea7m77rqLb775pk6PXVdOBzBc7PVeEH7Ofthpdnzy6yc88N0DbMjehGe1laJ9JRwaNprCVYc5ElnNS/8XwpzuCnsnO3qUlYNH6OUd0NUXDdDKTqFpGlpZ3pkEnkIIcQOSduTVJ+1IaUf+UTdMO1IpdcP+tGvXTp0vLS3tgnXX2pAhQ1Tbtm1V69at1XvvvWdb7+rqqiZPnqzatGmjevTooU6ePKmUUqp79+7qoYceUikpKSomJkZt2rRJKaXUk08+qe655x7Vu3dvNXr0aJWRkaF69Oih4uLiVI8ePdThw4dVQUGBio6OVnv27FFKKXX77ber999/XymlVHh4uMrJyVHp6emqRYsWauLEiSomJkaNGTNGrVy5UnXq1Ek1b97cdrxNmzaplJQUlZiYqFJSUtSePXtUZWWlCgsLU76+viohIUEtWLBAlZSUqD/96U8qKSlJJSYmqiVLllxwDUaNGqWWLVtmW77zzjvVokWLVHp6uurSpYtq06aNatOmjVq/fr1SSqn09HQVExOjlFJq5syZ6v7777dtO3DgQLV69WqllFLffvut6tixo2rTpo0aMWKEKi4uVkopNWXKFNWqVSsVFxen/v73v1/y/cnMzFRdu3ZVCQkJKiYmRq1du1ZNmTJFGQwGlZCQoMaMGXPR97G2cnPmzFHJyckqISFBTZo0SVkslguO+fTTT6ukpCQVExOj7rnnHmW1WpVSSm3evFnFxcWpjh07qocffth2DUpLS9XIkSNVXFycGjVqlGrfvr1KTU21fY6mTp2q2rdvr3788ceLHv/DDz9UUVFRqnv37uruu++2XdOlS5eq9u3bq8TERNWzZ0+VnZ2t0tPTVUBAgAoODlYJCQlq7dq16sknn1Qvv/yyUkqp7du3qw4dOqi4uDg1dOhQlZeXZ/vsPvrooyo5OVlFRUWptWvXXvLan83V1fWc5T59+qgNGzYopZQym83Kx8dHWa1WNW/ePDVp0iRbuUmTJql58+Zdct+fffaZiomJUfHx8apr165X9DmeOXOmGjx4sOrbt6+Kjo5WTz31VK3HWLNmjUpISFAJCQkqMTFRFRUVnfM5njhxou11X19f235eeukllZSUpOLi4tS0adN+8zpFR0erzMxMpZT+2Y2Ojr5o2bOPf6Wu5r1zd+5uFTsrVq3MWGlb9+xPz6qRS0eqkUtHqr+83UP91KO5SmvRUh0cdIsqWfy2Uk96KJX+o174i/uUeqXF5R8w+xd9+3+3VuqtDko9ZVJq2SN1fFZ1B9iiGsAzVX4azk9t7RtxbUg7UtqR0o6UdqS0I69cbffOi7Vv6v0hW58/vxXAmL5purrr67vq9Gf6pukXfeNOy83NVUopVVZWpmJiYtSpU6dOv4lq7ty5Sin9RnT6RnD6xqCUUj/88IPtg/Pkk0+qtm3bqrKyMqWUUoMGDVKzZs1SSin10UcfqSFDhiillFqxYoXq2LGjmj9/vurbt6+tHmc/eIxGo/r5559VdXW1atu2rfrTn/6krFarWrJkiW0/hYWFymw2K6WUWrlypRo2bJhS6sIHweOPP67mzJmjlFIqPz9fRUVFqZKSknOuweeff67uuOMOpZRSlZWVKjQ0VJWVlanS0lJVXl6ulFJq37596vR7eDkPnpycHNW1a1fbsaZPn66efvpplZubq6Kjo2038/z8fKWUUl9++aWaOnXqBe/PK6+8op577jmllFIWi0UVFRUppS68EV7sfTy7XFpamho0aJCqqqpSSin1l7/8Rc2ePfuCY57el1JKjRs3Ti1dulQppVRMTIzt4TtlyhTbNXj55ZdtN9tdu3Ypo9Foe/AAauHChZc8/vHjx1V4eLjKzc1VVVVVqkuXLrZrmpeXZ7tWH3zwgZo8ebJSSp3zoDl/OS4uTq1Zs0YppdTUqVPVQw89pJTSP7unt1+2bJnq2bOnUkqp48ePq/79+19wHc52/vWOiYlRR48etS03bdpU5eTkqJdfflk9++yztvXPPPPMOfWsTWxsrDp27JhS6szn4XI/xzNnzlSBgYHq1KlTtvf+9LU/26BBg9S6deuUUkoVFxcrs9lc640/IyNDtWjRQmVkZKhvv/3W1vCorq5WAwcOVD/88INSSqn+/fur48ePX3AcT0/Pc5a9vLwuet4NNYCx7cQ2FTsrVq3/ZZ5S2b/qP+WFylJcrLJfnK7SWrdWe2KjVe6M15TVbFYq/4gegJg7UqkV05R6LV6pD3pe/gEtVUr97/+UWjBO/1l4h1JHt1y18/ujJIAhP+f/SACj/kg7UtqR0o6UdqS0I6/clQQwJIlnA/TGG2/wxRdfAHD06FH279+Pj48PBoOB2267DYBx48YxbNgw2zajR48GoFu3bhQVFdnGHA4ePBhnZ33c+E8//cTnn38OwPjx43n00UcB6N27N//973+5//772blzZ611ioyMJC4uDoCYmBh69uyJpmnExcXZxogVFhZy5513sn//fjRNw2w217qvFStWsHTpUl555RUAKioqOHLkCK1atbKV6d+/Pw8++CCVlZV88803dOvWDWdnZwoLC/nrX//Kjh07MBqN7Nu377Kv68aNG0lLS6Nz584AVFVVkZKSgoeHB05OTtx9990MHDiQQYMG2a7d4MGDL9hPcnIyEyZMwGw2M3ToUBITE2s93sXex7N99913bN26leTkZADKy8vx9/e/YF+rV6/mpZdeoqysjLy8PGJiYujatSvFxcV06tQJgDFjxvDVV18BsG7dOh566CEAYmNjiY+Pt+3LaDQyfPjwSx5/8+bNdO/eHW9vfTrKkSNH2q71sWPHuO2228jKyqKqqorIyMhLXvfCwkIKCgro3r07AHfeeScjR460vX76c9yuXTvbZyk4OPiKu9np97lzaZp20fWX0rlzZ+666y5GjRp1zu/Z2S72OQb9d+r0ez1s2DDWrVtHUlLSBceYPHkyY8eOZdiwYYSGXji8oaKigpEjR/LWW28RHh7Om2++yYoVK2jTpg0AJSUl7N+/n27dujXq7rVl5jIAnJc+CJVVKAWFha05ucWO6txcvDpH4ee7FrsJfwY7O/AMhcA4OLRG/wGIvYIxoUZ7GPRqnZ+HEEJcC9KOlHbk+aQd+dukHXn9tCMlgHEJU9pPuebHXLNmDatWreKnn37CxcWFm266iYqKilrLnv3Lc/4v0ullV1fXix7rdBmr1cru3btxdnYmLy+v1l8AR0dH2/8NBoNt2WAw2MZFTp06lZtvvpkvvviCjIwMbrrpplqPq5Ri8eLFtGjR4qJ1c3Jy4qabbuLbb79l4cKFtgfrq6++SkBAADt37sRqteLk5HTBtnZ2dlitVtvy6eunlKJ3797Mnz//gm02b97Md999x4IFC3jrrbf4/vvvL1q3bt26sXbtWpYtW8b48eN55JFHLhijd7nvo1KKO++8kxdffPGix6uoqOC+++5jy5YthIWF8dRTT1FRUVHrDfXs/V6Mk5MTRqPxksc//cCszQMPPMDkyZMZPHgwa9as4amnnrpo2ctx+rNkNBp/1xjb00JDQzl69CihoaFYLBYKCwvx9vYmNDSUNWvW2ModO3bsop/N02bMmMGmTZtYtmwZiYmJ7Nix44IyF/scb9q0qdbfx7fffpsPPvgA0McFP/bYYwwcOJDly5fTsWNHVq1adcHn+d5772XYsGH06tXLdszHH3+cP//5z5d7WQgICCArK4ugoCCysrJqbdg0dLYcGGEdqfS7hezXZlKWXoBTTEvC3n0H5yOzYNcOsK+5fpoG966rvwoLIQTSjjybtCPPkHaktCNB2pF/hCTxbGAKCwsxmUy4uLiwZ88eNm7caHvNarXaErPMmzePLl262F5buHAhoEdMPT098fT0vGDfnTp1YsGCBQB8+umntu1fffVVWrVqxfz5820R4d9b95AQPVP/rFmzbOvd3d0pLi62Lfft25c333zTdnPcvn17rfu7/fbbmTlzJj/++CN9+/a1HSMoKAiDwcCcOXNqTR4UERHBjh07sFqtHD161Jaop2PHjqxfv54DBw4AUFZWxr59+ygpKaGwsJABAwbw2muv1XqTOdvhw4fx9/fnnnvuYeLEiWzbtg0Ae3t727W71Pt4drmePXuyaNEiTp48CeiJgg4fPnzO8U4/sHx9fSkpKbF9BkwmE+7u7rZ9n35vAbp06cJnn30GQFpaGrt27ar1XC52/Pbt2/PDDz+Qn5+PxWJh8eLFtm3Ofp/PzsR8/vt8mqenJyaTiR9//BGAOXPm2KLodWnw4MG2+ixatIgePXqgaRp9+/ZlxYoV5Ofnk5+fz4oVK2yfp8cff7zWh+zBgwfp0KEDzzzzDL6+vhw9evSKPscrV64kLy+P8vJylixZQufOnbn//vvZsWMHO3bsIDg4mIMHDxIXF8eUKVNISkpiz54959Th7bffpri4mMcee+ycY3788ceUlOjTdx4/ftz23l3OdZk9ezZDhgy57GvaUJSZS7E3K4ybDRx6+C0qTlkITC4gYgh68OLg9+AeVN/VFEKIeiftyDOkHamTduTlkXbkb1+XhtKOlABGA9OvXz8sFgvx8fFMnTqVjh072l5zdXXl119/pV27dnz//fdMmzbN9prJZKJTp07ce++9fPTRR7Xu+4033mDmzJnEx8czZ84cXn/9dfbt28eHH37Iv//9b7p27Uq3bt147rnnflfdH330UR5//HE6d+58zgPh5ptvJi0tjcTERBYuXMjUqVMxm83Ex8cTGxvL1KlTa91fnz59WLt2Lb169cLBwQGA++67j9mzZ9OxY0f27dtX6zcDnTt3tnVVfPjhh2nbti0Afn5+zJo1i9GjRxMfH0/Hjh3Zs2cPxcXFDBo0iPj4eLp3786rr+pdx5cuXXrONT5tzZo1JCYm0qZNGxYvXmzrYjdp0iTi4+MZO3bsJd/Hs8u1bt2a5557jj59+hAfH0/v3r3Jyso653heXl7cc889xMXFMXToUFs3PdAzJk+aNImUlBSUUrYGx3333UdOTg7x8fH861//Ij4+vtbGyMWOHxISwj/+8Q86dOhAr169aN26tW37p556ipEjR9K1a1d8fX1t+7rlllv44osvSExMtD1kTps9ezaPPPII8fHx7Nixo9brerbMzEwGDBhQ62uPPvoooaGhlJWVERoaaovcT5w4kdzcXJo3b85//vMfW5Zkb29vpk6dSnJyMsnJyUybNs3WpXHXrl0EBgZecIxHHnmEuLg4YmNj6datGwkJCVf0Oe7SpQvjx48nMTGR4cOHX9DtD+C1114jNjaWhIQEnJ2d6d+//zmvv/LKK+zatYvExEQSExOZMWMGffr0YcyYMaSkpBAXF8eIESNsD8MBAwaQmZl5wXEee+wxVq5cSVRUFCtXrrQ9yM6/xqNHjyYlJYW9e/cSGhp60ftIfTBu3M6/P6zGvGIfHv370eybbzD164KWvRN+/QIqCiGqV31XUwgh6p20I8+QdqRO2pHnknbk9d+O1C7VRaixS0pKUlu2bDln3e7du88ZQ9eQuLm52SJmZ7vpppt45ZVXav1wi8atpKTENo/19OnTycrK4vXXX6e6uhqz2YyTkxMHDx6kZ8+e7Nu3z/YAv5J9WywWbr31ViZMmMCtt956tU6lXvTt25dvv/22Tvc5a9YstmzZwltvvVWn+23orsa905KTw4kXp1O0fDmZJki8Zxg+E56v02M0BpqmbVVKyQNA2NTWvhHXhrQjxfVE2pF/jLQj605t986LtW8kB4YQ17Fly5bx4osvYrFYCA8Pt3W5LCsr4+abb8ZsNqOU4t13372ihw7oEfJVq1ZRUVFBnz59GDp06FU4g/pV1w8dUTeU1UrBZ59x8t//QVVUkD44ln+22E1qcu2JzoQQQghx5aQd+cdIO7J+SA+M66gHhhBCNFR1de+sOnyYrH9OpSw1FZeOHQl8chpvpb/NgsNfk9rldRkqUgvpgSHOJz0w6o+0I4UQ4spJDwwhhBDXFVVdTd7sT8h54w00e3uCnnsWz+HD0TSNsr3FOCsFTh71XU0hhPhNSqnfnGZRCCGE7ko7VEgAQwghRL2q3L+fzCf+ScXPP+PWoweBTz7JUccSVh9YAsC+kqO4WBU4SgBDCNGwOTk5kZubi4+PjwQxhBDiNyilyM3NrXVK44uRAIYQQoh6oaqqOPXBB5ya8R5GNzeC//0KHgMGkFeRx9gvxlJiPpNsLslikR4YQogGLzQ0lGPHjpGTk1PfVRFCiOuCk5MToaGhl11eAhhCCCGuufJffiXriSeo3LsXj4EDmdHxJEtPPQ6fPI5CYdQMzOk/B38Xf0j9CJ+1r0gPDCFEg2dvb09kZGR9V0MIIRotCWAIIYSoc99kfMOS/UtsyxGeEUxJngIWC6dmvMepGTOw8/Eh9J13cLv5JlbNjieuspIO5RUAtLUaSfSJAaM9WAGM4OBaPycjxFWiaVo/4HXACHyolJp+kXLJwEbgNqXUomtYRSGEEKJBkQCGEEKIOrdk/xK2n9xOc6/mFJuLWZ+5nlsMibi8+CEVv/6Kx+BbCHziCYyenmSWZFKiwWDveEYNnQN7lsHiiXBsC4SnQGUROLqDjCcXjYimaUbgbaA3cAxI1TRtqVIqrZZy/wJkvj4hhBA3PEN9V0BcaOvWrdx888225V9++YWUlJR6rJEQQlyZYnMxCX4JfDrwUz7pM4shmzWsd02m8MgBUh+4mc33pGD09ARg/yn977Uojwiwd4bmPUEzwKqn4Ku/wcHVkv9CNEbtgQNKqUNKqSpgATCklnIPAIuBk9eyckIIIURDJD0wLiH7hReo3L2nTvfp2Kolgf/4xyXLtGrVin379tmWp02bxrPPPlun9RBCiKuppKqEAJcAqo4do/Cxxxm7xcz2aHs+GeREvutWyjb8yL78ffg4+7Dj+E8ANDdF6xs7myD+NjiwCvIO6utaDKinMxHiqgkBjp61fAzocHYBTdNCgFuBHkDytauaEEII0TBJAKMBcnFxwcnJiYKCAg4dOkR+fj69evWq72oJIcRlK6ksJmHDSdInDwFNI+iFF2h561DGaBoWq4X7v7ufubvn2srHVlbi7tnkzA5unVEPtRbimqptTJQ6b/k1YIpSqvq3puTUNG0SMAmgSZMmlywrhBBCXK8kgHEJv9VT4mpq3bo1e/bs4dlnn+W5557jm2++4f777yclJYX169fz9ddf07Jly3qrnxBCXIwlN5eJc0/SZn8WTh06EPzC89iHhNhetzPYMaPXDKqsVfqKvV9j/9md4B5QTzUWol4cA8LOWg4FMs8rkwQsqAle+AIDNE2zKKWWnFcOpdT7wPsASUlJ5wdChBBCiEZBcmA0UDExMcycOROlFJ07dyYiIoI777yT6dOnM2LECAleCCEapJIf13Fo8BBiDlVz4I6uNJn58TnBi9M0TcPR6Kj/lOXpDyM3CWCIG0oqEKVpWqSmaQ7A7cDSswsopSKVUhFKqQhgEXBfbcELIYQQ4kYhPTAaqJiYGO68805SU1MB2LlzJwkJCWzbto2EhIR6rp0QQpxLKcWJF6eTN3s2dk0j+cfQAm4f2BXNcJE4eVke5KXr/8/aCWjg6n/N6itEfVNKWTRN+yv67CJG4GOl1K+apt1b87qMoxJCCCHOIwGMBmr8+PGMHz/etvzzzz8zadIkVq9eTXFxcT3WTAhxI6q2VlNRXWEboW80GHGycwLAWlFBdU4OebNnYxozBst9YzmybAhuq56B/03TNwiMhbGLwM4RlIJPBkP2rjMH8AgFozySxI1FKbUcWH7euloDF0qpu65FnWo5sExhLIQQosGQ1uJ14vnnnwfgrrvuqt+KCCFuSNml2RRUFpyzroV3CygowpyVjbJaCX3nHdx73MzevL0AuBnsIXYIVJXCznn6lKgBsVCSrQcvuj0Coe31nfk0u9anJIT4De9/9TTfZH1Oj9APaBvdlDZhJpwdjPVdLSGEEDcwCWAIIYS4gFVZqbBUoFAopSisKsTT0ROTk4kycxmnSk5iPnIUSkoxuLlh5+eHe2wsAMVVei8xN9+WMPAVfYfmMtjx6ZkDeDXRAxh2jtf61IQQl8moObPfwUrKrr9yxw9PoBntiQvxJDnSm/YR3iSFe+PpYl/f1RRCCHEDkQCGEEKIC+SU53Cq7NQ563ycfXAyOqEVleJ8SoEqwz4wEKOPD9qePbZyJeYSANydfc9sPHIWVBSeWXZwBaP84SNEQzZhwCP8b+FyNntlsSbgf8z1+zuph/P5eF067/1wCE2DFgHutI/0JjnCm/aR3gR4ONV3tYUQQjRiEsAQQghxgSpLFfZGe4JdgwE954VjtYGqoxlopaVY7MES6Iezp+8F25aU5wPgdvasIpoGzl7XpO5CiLqhaRp3tHuQJzc8SWbWUh5rHgd/eYgKczXbjxSQmpFHakYei7Ye45OfDgMQ7uOiBzMivEmO9CbCxwVNcmgIIYSoIxLAqIVSSh62QogbWpW1CkejI24ObiirFcupU1Tm5KBpGnZBQRxS2fjVdKBQSp2zbUlJFgBubsHXutpCiDo2sOlAXt/6Gp+EeNJ+5TTwCscpZigpzXxIaeYDgKXayq+ZRaRm5LE5PY/vdp9g0dZjAPi5O+rBjAgTyZHetAz0wGiQNpYQQojfRwIY53FyciI3NxcfHx8JYgghbljmajPOds5Ul5ZizsxEVVZi9PTELjAQg709xrwcLFYLSilyc3PJqsxizCdjAKhW1QC4eYTV5ykIIeqAo9G1H2bOAAAgAElEQVSR21rezrs73yUjtC0RX/wZPEMhNMlWxs5oICHMi4QwL+7u2hSrVXEwp4TNGXmkputBjWW79MCmu5MdSeEmWx6NuFBPHO0kMagQQojLIwGM84SGhnLs2DFycnLquypCCFEvrMrKyZJsyqtOcqrSDEYjRncXDCcL4KTeTTxHmck15lLoVIiTkxOvHnyVKFMUXUK6QO4BwnYuwslDemAI0RiMajGKj3Z9xNwWnfhnaR7Mvx3uXgWmiFrLGwwaUQHuRAW4M7ZDOADH8stsPTQ2p+exeq/eznK0M5AY5mXLo9E23ISbozRPhRBC1E6eEOext7cnMjKyvqshhBD1QinF3rnv4vKfN3GsMuA7YQK+f/kLhmX3wZ5l4BkG5nJedzGT6deMKSnTyC7NJqMkg+eixjDEIwZOZkNJKZydA+MG9svxQsZ9tIkKc7Vt3csjErglQQI84vrg6+zLwKYD+TL9ax4Y8TGec26FT0fBxBWXndsm1ORCqMmFW9uEApBbUklqRj6b0/U8Gm+vPoBVgdGgERPsQXKEd82PCR83ma1ICCGETgIYQgjRCB0uOsxrW1/DbDUD4GTnxD86/ANvJ+9ay//v4P+Y8eU/mfC1mdZHFZkhEP3Cy/h3GKAXOLkHmvWEMQug4Ahhc7qxvjSTe1fdC4A9Gl1WvQRWq17e3gVc/a76eV4PdmcVUVBmZkyHJrjXfLMc6etaz7US4sqMbz2eLw58wX/ztnP3bZ/CnFvhs/EwdjHYOVzx/nzcHOkXG0i/2EAASiotbDusBzQ2Z+QxZ+NhPlqXDkAzP1faR/rQPtJEcoQ3oSaXOj03IYQQ1w8JYAghRCO0IXMDq46sItoUjVVZOVBwgF7hvegX0e+CstaKCiremcn0r6tQzo6k3Z1Cad+OtIqtKVtthtwDEN1XX/YMY3KlAwOdW0OXyVBZhM+8MfgkjIU24/Uy7gG/64+axqiwXA8iTenbEk8XmTpWXJ+iTFF0Cu7EvN3zuHP4t9gPfgOW/AWW/Q0Gv6XPNPQHuDna0S3aj27ReuCz0lLNrmOFtjwaX+3MZP7mIwAEezrpOTRq8mg093eTvGVCCHGDkACGEEI0QqXmUgDmD5xPZXUlneZ3IqtmdpCzlaxbT/YzzxB75Ai7knwY9vqXxPv4nFso7xBYzeDfSl/WNJxDk0lM+wrSvjlTruNfICDmap3SVaOUYub6DE4WVwL632HD24bS3N+tTvZfVG5G0/TkhUJcz8a1Gsd9393HisMrGJg4Rr83rH0ZfJpDl7/V6bEc7YwkRXiTFOENN0G1VbEnu4jU9DxSM/LZcDCXL3dkAmBysSfprKlbY4I9sDca6rQ+QgghGgZpTQkhRCNUUlWCvcEeB6MDDkYH3OzdyCo9E8Cw5ORw4sXpFC1fjkNEBG/9yQ+PTp0ZdTp48fNnsOopUAqq9T/s8Wtx5gC9nj5nFgI8Qq7L4AXA8YJynvkqDTuDhsGgUWWxUlpp4ZkhsXWy/8JyM+6Odhhk6khxnesc0pkIjwg+3f0pA5sOhJufgNyD+r3Cuxm0HnzVjq3nxvAkJtiTuzpHopTicG6ZbchJakYeK9NOAODiYKRtE324SXKkiTZhJpwdZKYTIYRoDCSAIYQQjVCJuQQ3+zM9CILcgsgqyUJVV5O/cCE5r76GqqzE94G/4jnhLn78LIV73c+a9nTHPLBaIKqPvuzqC4HxZ173rftvXOtLfqk+xOOdsW3pExNIl399T0mlpc72X1hulqEjolEwaAbGtBrDC5teYGfOThL8EmDoO1B4FD6fpE+vGtL2mtRF0zQifF2J8HVlVLJ+7zpRVEFqzZCTTel5vPbdPpQCe6NGbIinbchJUri3/E4KIcR1SgIYQgjRCJVWleBqsIPDGwAINrqi9h8i470xVPz8My4pHTH98zHsmoSRXnwUhSLUXZ8dAEsVHN2k57MY8FI9nsW1kV9WBYDJVc/Z4epgR1ll9aU2uSIF5WY8neWPJdE4DGk2hDe3vcmnaZ+S0D0B7J3h9nnwYU99etV7vtcDGfUgwMOJQfHBDIrXZ/gpLDez9XAem9PzSc3I4+N16bz3wyE0DVoEuJMcUZNHI9KbAA+neqmzEEKIKyMBDCGEaIRK8vbjVnAcZvbHatbofcCXZj/bU2Uy4fn8k3wemcN760acs03YhhmwYRaYy8FcBhFd6qfy15gtgFHzjayzg5HSqrrtgeHlLAlNRePgYu/CrVG3Mm/3PLJLswl0DQQ3fxjzGXzUB+bdBhO+AUf3+q4qns729GgZQI+W+pTOFeZqth8p0HtpZOSxeNsx5mw8DEATb5eagIaJ9pE+RPi4SGJQIYRogCSAIYQQjVBpZRGuVivFLZ4m++3PiD6Vz8o2GvO6F1Fa8jzsgp5NepLolwiAW9qXJOzbAAE1eR+a94KmN9Vb/a+l07OEeLnU9MBwNFJWVXc9MArLzQR7OtfZ/oSob6Nbjmbu7rks3LuQh9o+pK/0bwUjZ8Kno2DRRBg9HwwNK++Ek72RlGY+pDTTc/1Yqq2kZRXpeTTS81i99ySLtx0DwNfN0TZta3KEN62CPDBKHhshhKh3EsAQQojrQLW1mvWZ6ym3lANgb7Cnc0hnHI2OtZZXhaXcutzIsX3v4RjVHK+bzQRGefJg8gQA3B3c6RfRD3tjzdCGH2ZAZHcY+9k1OZ+G5HQOjNPDPFwc7MgrLa+z/ReVm/GQISSiEQl1D+Wm0JtYtG8Rf47/M052NcMvmvfSh50t+zt8+wT0n16/Ff0NdkYD8aFexId6cXfXpiilOJhTYhtysjk9j+W7sgFwd7SjXYTJNuwkPtQTR7uGFaARQogbgQQwhBDiOpB6IpX7v7v/nHXPdHqGW6NuPWedUorCL5Zw/7v5OFVp+D74AL5334329WRGpS2FJkf0gvYuEFwCziaoLIacPRAz9FqdToOSX1aFu6OdbdpFVwcjZXU0hEQppSfxlACGaGTGtR7H90e/Z9mhZQyPHn7mheS79ZlJNr4DPs2g/T31V8krpGkazf3dae7vzpgOTQB9lqLU0zOdpOexZu9eABzsDCSGepFcM+SkbRMv3J3k91wIIa42CWAIIcR1oKCiAIA3bn6DEPcQRv5vJMdLjp9TpurIEbKefJKynzaSGapxqK8z/3ffffqL0f30mUXWvHhmgzUv6gn4rBZAQfC1mT2goSkoq8LL9cwfHi6OdpSelcRTKcXKtBO2oSYGTaNnK3/bkJPaZBWWU1pZTaWlGnO1kgCGaHSSApJoYWrB3N1zGRY17Nx8EX2eg7xD8PUUMEVCVK/6q+gfFOLlTEibEIa2CQEgr7TKNtNJakYeM344xNurD2LQoHWwh95DI8Kb5EhvfN1q7yEnhBDi95MAhhBCXAfKLGUAtPRuSZBbED5OPpwsOwmAsljIm/0JOW++iWY0EvjUk9xR/hyjjWemUaXVIJiWe2Y5cxvsXAiq5g91R3do2v1anU6Dkl9mxnRWMOL8Hhj7T5Ywac7Wc7Z5oEdz/t6nRa37+2L7MSZ/thOlzqzzc5c/ZETjomkaY1uNZdqGaWzO3kyHoA5nXjQYYfiH8HF/+O9dMHEFBLSut7rWJW9XB/rGBNI3JhCA0koL248UsDkjj83puczbdISZ6zMAaOrnqgczaoadhJqcJTGoEEL8QRLAEEKI60CpuRTQZwAACHAJ4GTZSSrS0sj651Qq0tJw69mTwGlTwdebirnP4+Zw3iwAZzecQ9rpP0LvgXFWAMPZwY6yqmqsVoXBoJFVWAHAu2PbEhviyR0fb2bfiWJb+YxTpYyY8RPlNUGPcnM1iWFe/KlzJAAORo2bWvhfwzMS4toY0HQAr217jbm7554bwAA9KDpmAXzQU5+Z5J7v9NlKGhlXRzu6RPnSJcoXgCqLlV3HC229NJbvymJB6lEAAj2caB+p985oH+FNlL8bBkkMKoQQV0QCGEIIcR04P4ARbOdD1KLtpK8bhdHbRMjrr+PepzeaplFQng+Aq6NHvdW3IaswV/PA/O2cLNIDE3uyi23fpoLeAwP0QISrox25JZUAtAzyIMzbheb+bhzKKbWV/3F/DqdKKrkjJRwHowEHOwN3pEQQ6Ol0Dc9KiGvP0ejIiOgRfPDzBxwtOkqYR9i5BTxD9dlIZg6A+aPhrq/0YWuNmIOdgXbhJtqFm7i3ezOsVsXeE8W2pKCb0nNZujMTAC8Xe5LCzyQGjQ3xtOXiEUIIUTsJYAghxHWgzFKGg8EBe4M9pRs3cvsLm3A7WYLXyBH4P/wwRk9PMksyqbBUcKLoMABujl71XOv68/OxAo7ln5lJxN/dES8XPQ/Ft7+eYGXaCTo188HBzkCnZj4MaxtiK+viqD8ay6r0AMapmgCGr5veS6OZnxs/7M3BUm3Fzmhgx9FCfN0ceXpwjHQPFzec21rcxse7PmbennlMaT/lwgIhbWHY+/DZeFhyHwz/CAw3zh/pBoNGqyAPWgV5cEdKBEopjuaV24acpGbks2q3PhzQ2d5ImyZetoBGmyZeuDhIU10IIc4md0UhhLgOlJnL8DU7kfnEExQu/hxjoBdPjzHw8eOPY7R3Yf6vn/DClpfP2cbk0vi6a1+O/NIqhr+7AXO1umiZtk28+PTuDrUGHE73wNDzYDiSW1KFg50Bt5rARlM/V6qqrazafQI/dye2Hs4jMcxTghfihuTv4k+fiD4sObCEv7b5K672rhcWaj0Yej0Nq57UZybp8c9rX9EGQtM0mvi40MTHhRHtQgE4WVzBlox8NtckBn3j+/0oBXYGjdgQT33YSYQ3yRGmSyYPFkKIG4EEMIQQogEot5TzUupLlFSVAGA0GLkn7h6aeTVDKYXf2p08uyCPwvLF+MSa2dw5j1+9Hei/uC9GzUh+RR4dyisYUaxv76gUKd2v38z/f8TKtBOYqxXvjW9HhI8rCsXx/HLKqs7MLNI+0vuiAYfT33ienokkp6QSPzdHW/lWgfrQnHvnbrNtc1tyk6tyLkJcD8a1Gsfy9OUsObCEsa3G1l6o80OQewDWvgzezSBx9LWtZAPm7+7EgLggBsQFAVBUYWbr4XzbTCez1mfw/tpDALQIcCc58sywkyDPxj0kRwghzicBDCGEaAD25u1l0b5FBLgE4GznTEZRBmHuYUwKGEb2M8/SbfUvHAtQxE9MwqlpKF2Ob2Zs4SEqND1g4aysTIwYhG+PJ/UdGh3A+cYZQjL+o038crwQgNKqasK8nenTOsAWdGgZePn5QFwdz+6BAbklVfi4nfnWMzbEg8V/6URJpf66UdNIijDVyXkIcT2K84sj3i+e+XvmM7rlaAxaLUNENA0G/gfyM2DpA2AKh/BO17yu1wMPJ3tubuHPzTXJfyvM1ew8WqDn0cjIZ8n2TOZuPAJAqMmZ9jXBjORIb5r6ukpvMCFEoyYBDCGEuAqUUizav4hT5acA0NAYEDmAJh61f1NfWa3nWZjedTpJgUncNL8bXl/9xKHPP0YpxZr+3myIzqH3uH+DewBeVWU8lvYlmPXpVXFwhdZDwf7GSxxZYa7mx/2naBduIiZYD1T0bBXwuxvxLjVDSEpremycKqkkwOPMddU0jXbhErAQ4mzjWo3j0bWPsu74OrqFdqu9kJ0D3DYHPuwNC8bC3av0ISXikpzsjXRo6kOHpj4AWKqt7Mkutg05+WFfDp9vPw7ouXqSws/MdNIqyB07SQwqhGhEJIAhhBBXwZHiIzzz0zPnrMstz+WJjk/UWv50AMPJzonKAwd4bFYJ4Ye34ty5M4FPP8WzP47HrcAKzjV/ODu4SBfsGqenOR3TvgnDa8aU/xGnh5D8uC+HonIzWYUVtA6SGV2EuJRe4b3wd/FnbtrciwcwQL+HjVkIH/aEeaP0IIazBASvhJ3RQGyIJ7EhnkzoEolSikOnSklNz2Nzhh7U+ObXbADcHO1oG26ifYQ+7CQhzAsne2M9n4EQQvx+EsAQQoir4FCBPl55Tv85xPvFM/DzgRSbiy9avsJSgZ1FYTfrcw7NXoS/veLL0eFMmfYBmqZRWl2Bv2bUv8EU58gq1GcbCaqjaUv93B0xGjQ+XJduWxfhW0tiQiGEjb3Bnttb3M4b29/gYMFBmnldomeFTzO4fR7MHgwLx8O4z+Xe9gdomkYzPzea+blxe3u9l19WYbmth0Zqej6vrNgHgIPRQEKYp54UNNKbduEmPJzs67P6QghxRSSAIYQQV0FGUQYAkZ6RGDQD7g7utgSdtdF27eVfM6tRp+bjMWgQ/73ZyvryX3isZhhEmdWMi+HGbWSaq60UlZtty57O9rZu0VkFeg+MIK+6SWbn6+bIT4/1oKhCP56maUT6SABDiN8yInoE7/38Hp/u/pRpKdMuXTi8Ewx+E5bcC8sm6/+X3A11JsjTmSGJIQxJ1KeILiir0mc6ychjc3oe7689xDtrDmLQ9BxB7SO9bbOd+Lk71nPthRDi4iSAIYQQV0F6YTo+Tj54OnoC4GrvSon5wgBGdUkJOf95lZD58znlDm5vTCekzxDctvyHU7vXoJTSe2AoMy7GG7dROeq9n9h+pMC2nNLUh/mTOgJnemAEetRd/g9/Dyf863B/QtwITE4mBjYdyP8O/o+H2j5ku/9dVOJoyDuoz0zi0xy6/N+1qegNyMvFgV6tA+jVOgDQkxTvOFJgG3KyMPUoszZkABDp60pyzZCTDpE+hHk7S2JQIUSDIQEMIYSoI2uOruFE6QkAtp/cTqRnpO01Nwc3skuzzylf/P1qsp95BsuJE+T0jefvMb+wasu9sPnP+Lq7UOXlSrG5GA8HD8qUFVe7G3e6vP0nSujUzId+sYFsPZzPlzsyOZJbRhMfF7IKKzC52OPsIOO6hahvY1uN5fP9n7N4/2ImxE747Q1u+oc+veqqp8C7KbQefNXrKPRcP52a+9KpuS+g93L75XihPtNJej4r0k7w2ZZjAAR4ONqmbU2O8KZFgDsGgwQ0hBD1QwIYQghRBwoqCnjg+wfOWdfTJw4Ofg+AW0UJxZVFAFhOnSL7+ecp/vobHKOiCH39NTYU/JfKjF9xjOoLnmH4ps0DYFPWJvxd/LFoei+OG1GlpZqSSgspTX24IyWCnq0C+HJHJqM/2IiHsz3H8ssINbnUdzWFEEC0KZr2ge2Zv2c+d7S+AzvDbzQ1DQYY+i4UHIXPJ4FnKIS0vTaVFTb2RgNtmpho08TEpG5gtSoO5JSwKT2P1JpcGl/9nAXoQ/iSwk0k1wQ04kI8cbCTmU6EENeGBDCEEKIOnCjTe1480+kZuoZ2haJsfN7tArwNgKuPiVIPbwoWf86Jl15ClZXh99CD+EyciObgQNXKjwFw7PI3CEog+MRWsB5l8prJtmOYHH6jO3YjVVCm56IwuepJ/kK8nJncO5pdxwsBCDU50y8msN7qJ4Q419hWY3lo9UN8f+R7+kT0+e0N7J1h9Hz4oCfMHw33fKcHMkS9MRg0ogPciQ5wZ3zHcJRSHMsv15OCZuSxKT2P7/acBMDJ3kBimBftI7xpH+lDmyZeuDrKnxhCiKtD7i5CCHERB/IPcKzkmG25tU9r/F38ay2bU54D6Ek7fZ19IX09oGDIO+DTDP/5d/PgVxVkHX4C56R2BD3zDI5Nm9q2rzCXYK8UBmdvABKCOvLx1m2U3fYJGOyxmz+a5HbDr97JNmB5pVUAeLuemaXgwZ5R9VUdIcRv6B7anRC3ED7d/enlBTAA3Pz16VU/6gPzbocJ34Cj29WtqLhsmqYR5u1CmLcLw9rqwaVTJZVsqRlykpqRx1urD2D9/gBGg0ZssIdtppOkcBM+bjduDichRN2SAIYQQlzEn779EwWVZxJHdgjqwId9Pqy1bE6ZHsDwc/HTV2RuB4M9qtVQ8ubOp9OcaiqN4DPtCfxuH4NmOLe7bWVVKU5WBU56LwstpC3J68ph1qiaEgpc/er2BK8T+WV6AMPkItMsCnE9MBqMjG01lpdSXyItN43WPq0vb8OA1jBqFnw6ChZP1KdaNUhum4bK182RfrFB9IsNAqC4wsy2IwWkpuexOSOPTzYetk1H3dTPlaRwE0nh3rSLMNHU11USgwohfhcJYAghRC0sVgsFlQWMjB7J8OjhfPLrJ6w+uppqazXGWhrUp3tg+KYtB4Md7F9JOdFkjRlPZdpuSlq68mivCj6/td8FwQuASks5DihwdNdXRPWGnk+CuUxfNthD3Mirdr71bduRfLZm5NuWY0I86NRMTy6XX6oPITm7B4YQomEb2nwob21/i093f8rzXZ6//A2b94L+/4LlD8OKf0K/F69eJUWdcneyp3u0H92j9WB7paWaXccK2XI4ny0Zeaw8KzGoj6sDbcNNelAjwkRsiCeOdhKsEkL8NglgCCFELcoseuAgwiOCGJ8YuoR0YXn6ctIL02luan5B+ZOFR/CsrsZx+cNYLRo5v7iTt88do4+RkDde52DFAvLzUympLNKHmJyn0lKBk9Lg9DdSdo7QdfIF5Rqrf3y+iz3ZxbZlLxd7tk/tjaZp5J3ugeFqX1/VE0JcIXcHd4Y2H8p/9/2Xv7X7W633vYtqfw/kHoSN+hA8ku++ehUVV42jnZGkCG+SIryhezOUUhzMKWVLRh5bDuez9XA+K9P0/FEOdgYSQj1pF64POWkXbrLlPRJCiLNJAEMIIWpRWlUK6NOfAsT6xgLw2rbXCPcIB6BXeC/a+LcBIKfoMH7V1ZRGPU7WxyswH8/Ca+QI/B95GKOHB26rvod8KC3OAq+mFxyvwlKBo3bjZnHPLqpgdPswnhjYmi+2HWPql7+SfqqUpn5u5JfKEBIhrkdjWo1h/p75fLb3M+5LvO/KNu77POQdguWPgilC75khrmuaptHc343m/m7c3r4JADnFlWw9nM/Ww3pQ46N1h5jxgwKgub+bLZiRFOFNhI+LDDsRQkgAQwghalNiLgHOTF0a7hFOrFsTUjM3kJq5AbOysmjvZ9wVNwEDBo5kHWDs13AkbTYO4eE0mT0b1w7tbftzc9dnySguOgJ0vuB4VdZKHLUbs/tshbmagjIzwZ7OuDna0aGpDwDbjhTQ1M+NvNIq3B3tsDfeuAEeIa5H4R7hdA3tysK9C7k77m4cjFcQhDQYYcRH8HE/+O+fYOIK8G919Sor6oWfuyP9YgPpF6s/IyvM1ew8WmDrofH1L9ksSD0KgK+bgx7MCPcmKcJETLBM3yrEjUgCGEIIUYtSs94D43QAw1CUxfxfNoCyApBtNHJXWBPe2f42nXYrHl1pxaNCw2fS3fjedz8GJ6dz9ufmrmdt/+6XORw/9hMGDHRrOwkfPz25XUW1GUfDjXlLzimuBCDAQ79mzf3ccHe047llabyz+gA5xZXSlViI69TYVmP588o/803GNwxuNvjKNnZ012cm+aCHntjznu/02UpEo+Vkb6RDUx9bINtqVRzIKWFLRj5bDuex9XA+3/6qDztxtDOQEOZFUriJ5Ahv2jYx4ekiQw2FaOxuzNayEEL8htMBDLejqZB/Ag58rwcv7vke3IMI3PguX658m+y8IZSt34hjiCvBnc04Tf57rfsLCGyDg1IsqDgK2fq3SXet2svfR38NQKWy4GhwuTYn18CcLK4AwN9Dn2bPYNB4fEArfjqUayvTtfkVjJ8XQjQYKUEpNPNsxty0udzS9JYrHwLgGQqj58PMgbBgDNz5P7B3vjqVFQ2OwaARHeBOdIA7Yzrow05OFlewNSNfTw56OJ/31x7inTUHAYgOcLPl0UiO8CbM21mGnQjRyEgAQwghamEbQvLd82DWZ8GgWU8IaYeqrib/gBs5X/uh2EhAVydMTY6jRXa66P5M3s1YM+J7SivyAJi4fCzHi4+AUqBpVFoteBgdr/p5NUQnivQeGP7uZ3qtjOnQxNZYFUJcvzRNY0yrMTy78Vm2n9xO24C2V76TkHYw7H34bDws+QsM/xhqmc1J3Bj83Z3oHxdE/zh9+tbyqmp2HC2w5dH46udM5m8+AuhDVE7n0UiO8KZ1sIcMRxTiOicBDCGEqEVpuR5ocGszHpLv1Vd6NaFy/36ypk6jfMcOXFsGEdjPHweTIxALbe+45D7d3fxxr+n+HOwazInyvXqD3M6ZSlWN4w0awDhZpPfACPC4Mc9fiMbulma38Pq215m7e+7vC2AAtB4MvZ+BldPAMwz6PFu3lRTXLWcHIynNfEhpdmbYyb6Txfqwk5oZT77+JVsva28kIczTlkejbbgJDycZdiLE9UQCGEIIUYvS0hwAXPxag18LrFVV5L77Hqc++ACjqyvBL/0Lj1t+R3foGgG+Lfmp+DBk/wJApYc9Tu7BdVb/hqaowszBkyW25YM5pRzJ06eq3XgoFzuDJrOMCNFIOds5Mzx6OJ/8+glZJVkEuQX9vh11ehAKjsCGN8CriT7dqhDnMRg0WgZ60DLQg3Ed9VnDThRVsCUjn9QMPY/Guz8cpHq1QtOgRYC7rYdGu3AToSYZdiJEQyYBDCFEo5WancqW7C225RjfGLqFdrusbUvKTwHg6h5E2bZtZE2dRtXBg3gMvoWAxx7Dztv7D9Ut0COcUwYNywNbsDPYUbHwJhy8wv/QPhuSbUfyWZWmJ1pTwOfbjtmGitSmQ6Q3BoM0GIVorEa3GM0nv37C/L3zmdxu8u/biaZBv39B4XH4+lHwCIGWA+q2oqJRCvBwYmB8EAPj9eBZaaWFnUcLSK1JDvrljkw+3XSkpqwjSeHetqBGqyB37GTYiRANhgQwhBCN1supL7M7b7dt2c3ejXW3r8No+O3pSksr8jGVV5Pz0dcU/O977IODCfvgfdy6dq2TugW4BmBVVk6VnyLQNZCq6iqc7Jx+e8PrxMvf7OWnQ7nYG/WgRJjJhXfGxuDsoF97X1dHYkM85FsuIW4QQW5B9GjSg8X7FnNv/L242P/OpMVGO3161VmDYNEEuGsZhLar28qKRs/V0Y5OzVjgwhQAACAASURBVH3pVJMgutqq2JNdxNbD+WzJ0KdwXbYrCwAXByOJNbOdJEV406aJF+4y7ESIeiMBDCFEo1VuKadvRF9e7vYyy9KX8fiPj7M3fy+tfVr/5rYeWzOYvkhRULoa7zvvwO/BBzG4utZZ3QJd9DnvV2SsIMIzgvLq8kaVAyP9VCnD2obwn1GJ9V0VIUQDMa7VOFYeXslXh75iVItRv39HDq769Kof9oJ5o+DuleDdtO4qKm44RoNGTLAnMcGe3JESAUBmQTlbDueztSaPxlurD2BVYNCgRaBHTUBDD2qEeMnMOEJcKxLAEEI0WlWWchzL8tHSfyCpUp9JZPmh5VRW18x64eJPiFvIOdtYcnLIfv4Fun5zgCxfRcS7H+Kc3LnO6xbhEQHAy1tetq3zdvpjw1IaitJKC9lFFTT1rbuAjxDi+tfGvw2tvFsxb/c8RkaP/GM9sNz8Ydxi+Kg3zB0BE1eCq0/dVVbc8IK9nBns5czgBD0/VUmlhe1HzvTQWLztGHM2HgYgyNPpnDwarYI8MMqwSCGuCglgCCGuGwcLDrL++HrbcoRnxCVzWlSW5uCYfQA2LSIQiGgSxuy02cxOm20r42xX862JUgzfZ6L/VycwmKv59iZHtiSU0yPp4lOj/hFhHmF8detXFFcVA2DQDESboq/Ksa61jNxSAJr6udVzTYRo2DRN6we8DhiBD5VS0897fQjwLGAFLMD/KaXWXfOK1hFN0xjXehxPrHuCn7J+olPwH7y/+kbB6AUwezAsGA13fAn28k24uDrcHO34f/buOzyqMnvg+PedmWQmkzLpIZVOKAkdUVEEBcGCilRFVlF0bbv29bdrVyxrW7tYQBRRECygYkVdQUTpvUkRQgkkpE8yycy8vz/uZBIgUViSTBLO53l4nLnzzr3vBZO5c+55zzmzfRxnto8DwO3xsml/kb/TybKdeXy2xlh2EhpspkeaL0OjZTTd0yIJs8rXLiHqgvwkCSGajBdXvsiCXQv8zy0mC8uvWI5J1Vxcqxwv1ogUuOg9+O1bXl/8HDtGvgaOZNCwNX8rB50HsWcXkjllIbEbd7MxTfHm+UHsi/IyotRsFI2rJy0jmk/RztJyDztyjMDFz9tyAWgtGRhC1EopZQZeBgYDWcBSpdQ8rfWGasMWAPO01lop1RX4AOjY8LOtO0NbDeXZZc8yY+OMEw9gAKSdCpe+DrOvgo+ug1Fvg0kKLor6ZzGbyEh2kJHs4Kp+rdFasye/1F9HY+nOQzy/YCvat+ykU2KEP0Ojd6soEh0SbBPifyEBDCFEk1FSUUJmbCavDX6N9ze9z4srX8RZ4SQsuOY7/S40wbZIaHk6hCWQuPAZEg/sgNiuAJyW1pHcmR+S89K7qOBg4q+9gI6dLFxqUrDlSwgKb8jTa9JunbWSr9Zn+58HmRWtYiSAIcQfOAX4TWu9HUApNRO4GPAHMLTWxdXGh2I09WnSgs3BjE4fzaurX+X3wt/rJpDb5RIomARf3wNf3wtDHzvxfQpxnJRSpETZSYmyc3F3Y3lqYVkFK3fls3znIZbuzGPW0t1MW7wTgOTIEHq2jKJXWiQ9fctOgqTbiRB/SgIYQogmw+VxYbfYCQ8OJzbEqBxeVF5UYwBDa025UlgrO3tEtwFHGnz/KHz/KKWHgti3IgFXjpfwwYNJuPMmgqadBiuCwBxsvKfHFQ10Zk1PVp6TNVkFgFHz4usN2YzslcKgTgkAJEXa/B1HhBA1SgZ2V3ueBfQ9cpBSajjwOBAPXNAwU6tfo9NH88baN3hv43v8s+8/62anp90EBbthycsQmQqn3lA3+xXiBETYgjirQxxndTCWnVR4vGzcV8gyX/vWpTsO8enqvQDYgkx0TY6kR8tIeqZF0TMtirjw5lPcW4i6IgEMIUST4fK4CA82siIq/1tYXkgiiUePLTeWMwRbfK36lIIr5uD9fQUHZy3g0Dc/YbFWkPyPK4m45HLY9DloD1y9AJJ6NMwJNSHvLvmdp77ajNbGDeBilxtvtXvBVouJO89Np4Wj+bSCFaKe1bQ+7agMC631x8DHSqn+GPUwBtW4M6WuA64DSEtLq8Np1r3YkFjOb30+n/z2CTf3uNn/+/yEKAVDHoOCLPjynxCRDJ0vOvH9ClGHgswmuqZE0jUlkqvPaA0Y3U6W/57Hil15rNiVz9RFO3jNsx2A1OgQeqZF0aulEdDo2CIci2RpiJOcBDCEEE2Gy+3ytxqtvOCtLIJ51NiyQwBYqxV0K96Sw/4HplKRlUXkyOHEB72LedeT8MKTxoDIlpAobT9r8sW6fYQEmRmaYbR/jQgJYnCnBIItxoVUpD2IhAgJXghxHLKA1GrPU4C9tQ3WWv+olGqrlIrVWufU8PrrwOsAvXv3bvRLTcZ1Gse8bfP45LdPGN95fN3s1GSGS9+Ady6Cj66F8BaQekrd7FuIepIUGUJSZAjDfN1Oyio8rNtTYAQ0fs/n5225zF1l/GoICTLTNcVBT19Ao2daJDFhkqUhTi4SwBBCNBmu8iKsB7fAd48SXm4EKGoLYJSX5gFgDQrFnZfHgSf+TcHcuQS3akXL6e9g79MH9o6CA5uq3pTUvV6LdjZVWmvWZhVwYbckHryoS6CnI0RzsRRor5RqDewBxgKXVx+glGoHbPMV8ewJBAO5DT7TetA5pjM943syY+MMLut4GRZTHV2SBtuNziRTBsN7Y2DitxDTtm72LUQDsAWZ6d0qmt6tjNbqlcVBV+zKZ4UvU+ONH7fj9qVBtoqx0zMtih4tjYBGeoJkaYjmTQIYQogmw1V6CGtRPqxfRITFDKlJFPkyLY4emwdaE7Mql+0PXIinsJCY6/9K7A03YLL67lYk9Thpl4tk5TmZvuR33B7jAijUauHqfq2ItAcfNXbXISeFZW4ykx0NPU0hmi2ttVspdTPwFUYb1ala6/VKqet9r08GRgB/UUpVAKXAGF25jqsZ+EuXv3Dr97fy7a5vGdpqaN3tODQWxs0xghjvjjCCGKGxdbd/IRpQ9eKgF/myNErLPaz1Z2nk8ePWHD5auQcAe7CZbimR9PTV0uiRFkV06NGf7UI0VRLAEEI0GS7twRqeCDfvIHzZFFj/HEWFWTWPzdrJ/832krTtV4IyM0l7ayq29PQGnnHjUe72snhbDhW+gMULC7ayYV8hIUFGoU1nuZvJP2zDbFJ4tfb9AU+1QhcSwBCibmmt5wPzj9g2udrjfwP/buh5NZSBqQNpFdGKt9a9xZCWQ1B1mQEX09bIxHh7mJGJceWnRnaGEM1ASLCZU1pHc0rrqiyNrLxSf0Bj+a48Jv93u/8zvE1sKD3SovxBjQ4J4ZhNknEqmiYJYAghmgwXGqvZqLMQFtMegMKiPYeN0R4Pee+9T/nTz9LZqym4oj8d//kKynxydcRwuT0s3ZFHhdcLwIwlu/h2Y/ZhY166vAcXdjXu5qzNKmDeauPv0mRSmJTCpMCsFEopYsOtdEmKaNiTEEI0ayZl4souV/LQzw+xdP9STkms43oVqafAiDdh1niYMwHGvAvmoLo9hhCNgFKK1Gg7qdFVLVyd5W7WZFXV0vhh8wE+XGHc9AmzWuiW6vB3O+mRFlljBqYQjZEEMIQQTYLWGpfCH8CwRLXG7vVSVHLAP6Zsyxb23XcfZavX4M1M5Y4z9vLIOf1PiuDF3vxSnv92qz9gsXFfERv3FR425tZB7f1tTiNsQaTFVN2NzExxkJkiGRZCiIY1rO0wXlr5ElPXT637AAZAp2FwwdPw+R0w92a45FUwSX0A0fzZgy2c2iaGU9vEAMZ11K5DTn9AY8WuPF7+/jd/R7G2caFGQMNXILR9fBgmydIQjZAEMIQQTUK5txwAa5Cv00VEEuFeTVFpLl6Xi9zXXiPnjTcxh4WR9NRTbHas4uCOWQRbT44v5a/9dxtzVmSR6GtjarWYeHJkV9rHhwHG3Zb2CXXQqlAIIeqQ1WxlXKdxvLDyBbbkbaFDVIe6P0ifieDMg+8nQUgUDH1cCjaLk45SipYxobSMCWV4jxQASlxuVmfls9JXIPTbjdnMXm5kaYRbLXRPizSWnvj+6wiRDCYReH8awFBK/QA8rrX+qtq2W4EOWusbaxl/p9Z6WR3OUwhxknO5ywCwWnxZAyYz4cpM6Ob97LjwPMp378Nx3mDi738IS1QUrm9/MsbbIgM15Xq1N7+UR+dvpMJtZFws+i2HYV0TeW7syVmUVDQvSikPsBbjOmUHMF5rna+UauV7PklrfZ9vbCywD3hNa32zUiodeA2IBKzAQuA2YBfQRmtdUO04nwDvAXbgKYxuIJUu11pvqM/zFIbR6aN5Y+0bvL3+bR4949H6OUj/O8GZC7+8CvYYOOuu+jmOEE1IqNXC6W1jOb2tUeRWa83OXKe/28mKXfm89N1Wf5ZG+/gwX5aGUUujbZxkaYiGdywZGO9jtPb6qtq2sYD85hdCNBiXy1gOYQ0KBcBTVMTIbxQ9VxaTE17Egks82FO/5NYwo96dq8Jor9pcAxgvfreVr9fvp22ckWHRLj6M6/pLq0DRbJRqrbsDKKXeBm4CKr/ZbgcuBO7zPR8FrK/23heA/2it5/ren6m1LlFKfQ1cArzt2+4AzsBoXToamKW1vrlez0rUyGF1MKL9CGZumsnfevyNFqEt6v4gSsGQx6DUl4lhjzIyM4QQfkopWseG0jo2lBG9jCyNYpeb1burWrh+tWE/s5btBiDCZqG7L0OjZ1oU3dMiibBJloaoX8cSwJgDTFJKWbXWLt/djyTgcqXUf4AQYI7W+oEj36iUKtZah/kejwQu1FpfpZSKAyYDab6ht2qtfzrx0xFCNFeuUqNdqjU4lKIFC9j/0MP0OOhmST8H/z0nmix9iJyKYv62/mPMnS/B5TICGMHBoYGcdp35eVsuN8xYTnGZGwC3VzOubxqPDs8M8MyEqHc/A12rPS8FNiqlevuyPccAH2BcmwAkAv72RFrrtb6H7wM34AtgAMOBL7XWzjrtfiH+J+M7j+f9Te/z7oZ3ubPPnfVzEJMJLn4Jygrg8zvBFgmZI+vnWEI0E2FWC/3axdKvXVWWxvacEl9AI5+Vu/J4fsFWtDbihO3jw+jVMooeqUZxUMnSEHXtTwMYWutcpdSvwFBgLkb2xSyMZSWHlFJmYIFSqqvWes0xHvd5jLsji5RSaRjZHZ3+t1MQQpwMXKV5RBZrkt9aS9aKL7Cmp5Py8kt0zsxkAvDehhk8vvQJ8ubdROwnN1AeFgpxMVjN1kBP/X+SlefkXx+vo8Bp1P7YmevEERLEuL5G3NdiMjH+tJaBnKIQ9c53jXEOMOWIl2YCY5VS+wEPsJeqAMZ/gO+UUouBr4G3tNb5wJfAm0qpGK11Lsb1zIvV9jlGKXVGteenaa1L6/ykRI2SwpIY0moIs7fM5rpu1xERXE9dj8xBMOoteHcEfPxXI4jRflD9HEuIZkgpRdu4MNrGhTGqdyoARWUVrN5dwHJflsbna/bx/q9Glka41ULXVAfdUyPpnhpF99RI4sKb5rWZaByOtYhn5TKSygDG1cBopdR1vn0kAp2BYw1gDAI6V7vjEaGUCtdaFx3rxIUQTUBJLmyeD9qo00BIlFER/jjvdmqtcX36Nf9504Pds4+4224j5uoJqKCqNMW40HgAcvvdTKw1BlfBBsj5OWABDI9X89ZPO8jKq/r+M6p3Cl2Sai4qWlbhYcYvuygsrQDgx60H2by/yN/jPSHCxi2D2tf6fiGamRCl1CqgFbAc+OaI178EHgGyMW6q+Gmt31JKfYVx4+Vi4K9KqW6+LNJ5wEil1IdAd4wARyVZQhJgEzImMH/HfGZvns01mdfU34GCQuCy92HaBTDrCvjLXEjrW3/HE6KZC7cFcUb7WM5ob2RpeL1Glsaq3fms2p3Hqt35TP7vdjy+YhopUSG+gEYkPdIi6ZLkwBbU/DvGibpxrAGMT4BnlVI9MZaM5AF3An201nlKqWmArYb36WqPq79uQu5sCNHsuL1uNuZuxKM9ANiXv02HX6YePuimpRB37FXmy3fuZN/9D6B+/ZWdaZByy3g6X3DdUeNiQ4wPzYMdh5CefAau9W83eACjoLSCYpexxOOb9fuZ9PlGwm0WTEpRWuHhszX7uOJUI4PCq+FgkQu3xwju/HawmJW78v37CjIrJl2SwZg+aUcfSIjmr1Rr3d1Xp+IzjBoYL1S+qLUuV0otB+4AugDDqr9Za70XmApMVUqtAzIwAiHvA/cCCpirta5oiJMRx6ZjdEdOSzyNGRtnML7zeILNwfV3MJsDrvgIpg6B90bBhC8goUv9HU+Ik4jJpGgXH0a7+DBG+mpplJZ7WLe3gFW78lm12+h88tmafQBYTIrOSRH+oEb31Ehax4Yiy/tETY4pgKG1LvZ1F5mK8eEfAZQABUqpBOA84Ica3pqtlOoEbMZYa1qZYfE1cDNGxW+UUt211qv+99MQQjQGs7fM5rFfHjts2/T4dnQfNw92L4E5V4MzB/jzAIauqCB36lvkvPwyymrFOeEsHk5YxNRWrWoc7w9gOA8CUO7xtV2twwBGhcdLaYXH/zzcavF/uB4oKmPAUz/gLK96vVtqJJ/ceDpKKTbvL2Lcm0t47tut/tdjQoOxWkyA8WH/6PAMxvWVZSFCVNJaFyil/g7MVUq9esTLzwD/9S119W9USg0FFmitK5RSLYAYqrqLfI9RA+Mm4G/1fgLiuF2VcRV//eavfL79c4a3H16/BwuLh/GfGEGM6ZfC1V9CdOv6PaYQJ6mQYDN9WkXTp1W0f9uBwjJW7jYCGqt25fPh8ize+fl3ABwhQXRLjaRHaiTd0yLpnhJJVGg9BjVFk3GsGRhgBC4+AsZqrTcppVZiVP3eDtRWgPP/MO6c7AbWAWG+7X8HXlZKrfHN4Ufg+uOfvhCiMckrywNg8qDJeLWX2769kfkRDro7kqGkjTGoNP8P9uAbsnYt++69D9fmzYSfey4J997D4k2voXcobCE1dxWpDGDkluUC4PK4UCgspuP5NVe7sgoPA576gf2FZf5tp7eN4e/ntEcBn6/dh7Pcw4PDOmMPtoCCQZ0S/AGO9BbhLL3n8HXWcmdBiD+ntV6plFqNsYR1YbXt6zm8+0ilc4HnlVKVP6x3aa33+97j9S0fGYVx7VHdkTUwbtRaL66r8xDH5rTE0+gY3ZGp66ZyUduLMJvqOa08qiWM/xjeOg+mDzeCGOH10AVFCHGU+AgbQ7q0YEgX42fO49X8dqDYv+xk5a58XqzWxrVVjL0qSyMtis6JEQT7bgSJk8cxX9lrrT/GSLmsfH5VLeMGVHs8B6OLyZFjcjCqhgshmhGXx0WQKYh+yf2gooz+TidzQnP5fs5g8LhxJLVgakk2tZVm8zqdHHzhRQ698w6W2FhSXnqR8AFnGvt2FQAQbIuq8b0hlhDCgsIOy8Cwmq11FiT4dmM2+wvLuK5/G+LDreQ5y3n1h20s3pbrHzMgPY6r+tV+904CFkIcm8oOZtWeV18iklHD+GnANN/j24Hb/2DftwC31PZ+EVhKKSZmTuTO/97JN7u+YWirofV/0PhOMG4OvH0RvHMJTJgP9ug/f58Qok6ZTYr0FuGktwj3L6EtcblZk1Xgr6fx8/ZcPlm1F4Bgs4kuyVVLT3qkRpEaHSLXW81c3dyaFEIIoLysABsKlkwGZw5/zSsgPK0f3qiWHCjaw+KyHHYW/n5YP8RKxQsXsf/BB6nYs4fIsWOIv+MOzEuegkkjAHCF2SEuFltI7ReVsSGxLM1eyuTVk1l+YPkJrZ8uLfcw8Z2lbNpnrHwrKXeT6LBx99COmH3twC7tmUJ2QVVGhhTYFEKIEzcobRCtHa15fc3rnNvyXEyqAe6wpvQ2CnvOGGVkYlw5z6iTIYQIqFCrhdPaxnBa2xj/tn0FpazcVbX05P1fd/HWTzsBiA4NpluKg26pkXRLiaRrioOYMOl60pxIAEMIUWdcB9YRXF4KX94NQLopiIdOfxCi27B6/woWf3UlBWW5h73HnZfHgSeeoGDuPIJbt6blu9Ox9+5tvPjbAojrBF1H4SrcAjmLCQ4Oozbd4roxd9tctuYZdSa6xfZg3R4jc0NrWLApm+0HS/zj+7aJ9tec8Ho1D8xbz5osY4lLQWkFO3OdjO6d4k9PPKdTgj94AfjbiAkhhKg7ZpOZazOv5V+L/sUPu3/g7LSzG+bAbc6CMdNh5uXw3hi44kMIDm2YYwshjlmiI4TEzBDOz0wEwO3xsjm7iFW781m9O5/Vuwv4YctWtG/pSWp0CN1SjIBGt9RIMpIjjOW+oklq8H85pVTxkamhQojmweUuw4qGf+wwNlhsEGwHwOHLnMgvMwIEWmsKP/uc7Mcew1NURMwN1xN7/fWYrL4oeUUZHNwE/W6BM+/AtXEG5CzGZq6p4ZHhkX6P8NDpD/n3f94Li7hw4aLDxrSKsaOUoqzCw7zVe1m8LRd7kJncknK+23SAU1pHYw82ExUazLX920hRTSEaEaXUg0Cx1vrpABy7FbARozA5wBKt9fW+13phLEEJAeYDt2it9dF7EcfqvNbn8cqqV3h9zesMTB3YcCnhHYbApW/Ah9fAzHFw+SywyN1bIRozi9lElyQHXZIc/uu2EpebdXsKWJ1lBDSqdz0xKeiQEE731Ei6pkTSLdVBekI4FrPU02gKJPQkhKgz5W4Xwaga1w5HWo3imwWuAir27GHfQw9R8uNCbOmtSHvkemytkuHgWkjuBUrBgQ2stZj4qmIvLH2adbnrAPh09QGCTYUAxIRa6dcuFpPv82b7wRJ2HXICsCvXydbsEm4e2I5uqcax06LtpLcIB4yOIrfOXMXK3/P8cxzXN41Jl2TI2kkhToBSyqy19vz5yMZNKWXRWruP2LxNa929huGvAtcBSzACGEOBL45z36Iai8nCxMyJPPjzgyzeu9iordRQMi6FilKYeyPMngCj3wZzUMMdXwhxwkKtFvq2iaFvm6qlJznFLtZk5bNqdwGrd+fz5fr9zFy6GwBbkImMJIc/oNE9NZK0aLtcEzZCxxzAUEqZgSlAb0ADU7XW/1FKtQMmA3GAB6OydzYwF4gCgoB7tdZza9jnXcBowAp8rLV+4MRORwgRSGXecqzUXDE+PDgck0cT/lM22568CICESzoTFfwtatFiqEyUCIkCUxC4XbwRGcEP2Yux5a4AwOJO4b6PNwPHFiGPC7dy89ntsAUdPacgs4mXx/U87nMUojnxtSbtg5E5MEdr/YBS6jxggtZ6tG/MAOAOrfUwpdS5wEMYn9vbfOOKlVI7MVqtnwu8pJQKx/hCHwz8BozXWjuVUm2BGYAZ4wv+7ZVZmbVdEyil7gH+gtHR7CCwvIbzaOk7fpxvzASgAFgNtPF1H7FjZE+0AdKAl33jncC1vg5r04BDQA9gBXDHMfwdJgIRWuuffc/fAS7hiACGL3skCWgF5CilvgZ6a61v9r3+GfC01voHpVQx8DxwIVAKXKy1zlZKjQIewLjeKtBa9/+z+TVlF7W9iMlrJvPamtc4Pen0hv0i0WMcVDhh/p3wyQ0w/DWo744oQoh6FRtm5eyOCZzdMQEwsnV3HyplVVbl0pN83vv1d6b+5AUg0h5E15RIuvtqanRNiSQuXDKyAu14MjC6A8la6wwApVRlL8MZwBNa64+VUjaMbxblwHCtdaFSKhZYopSaVz2d0ncR1B44BaO7yTylVH+t9ZFtzSrHD8X4MDcDb2qtnzjideV7/XyMi5GrtNYrjuP8hBAnqNxbjrVasTWtNZuzi3B7NGz/jceme2mzrxB7/zNJfOABguaPp9DbnbIB9wOwZ/M6SncsQ2uNtsBm23YctCTdfTvFLjdLth/i0eEZDEyPB2DD3kI2Zxf5jxcTGkxGsoPKa9yECFuNwQshhN89WutDvpsUC5RSXYFvgNeUUqFa6xKMrmGzfJ/n9wKDtNYlSqm7Mbp9POzbV5nW+gwApVSM1voN3+NJwDXAixif089rrd9XSvnbp9d2TQCUYLRP7YFxzbKCGgIYwEvAO1rrt5VSVwMvaK0v8bVfPQv4HhgGfKW1rlBKvQ5cr7XeqpTqC7wCVBZa6OA7x5qySFr72sgXYtycWQgkA1nVxmT5ttWkF3CG1rpUKXVVLWMAQjGWqNyjlHoSuBaYBNwPDNFa76l2HdZsBZmDuDrjah775TGWZS+jT4s+DTuBU66F8mL49kEICoFhL4DcjRWi2VBKkRZjJy3GzkXdkgAjQ3dLdhFrsowsjVW783np+4P+Vq7JkSF0S3X462lkJjsItcqihoZ0PH/b24E2SqkXgc+Br313WJJ9LVbRWpcBKKWCgMd8Fx9ejA/yBGB/tf2d6/uz0vc8DOPi5agAhu/C6mVgMMaFwVJfQGRDtWHn+d7fHuiLkc7Z9zjOTwhxglxeN8GqKmDwr4/XMefnbVy2+VtGbf2eOJvmu/Mt3PjMayjt5dWSHbwSFQZLbgVAe2wEF91HWHA4ZpPikPoXtrII9pSVYTEpLshMZEzvVP8axaTIEAZ1TgjIuQrRTIxWSl2HcT2QCHTWWq9RSn0JDFNKzQEuAP6BEQjoDPzkuxMeDPxcbV+zqj3O8AUuIjE+37/ybT8NIzsB4D2gspZFbdcE4RjZGE4ApdS8Ws7jNOBS3+PpwJPV5jQGI4AxFnhFKRUGnA7MrnZHv/ottdm1BC/2AWla61xfzYtPlFJdqNZivpra6l/M01qX1vJadeXAZ77HyzGufwB+AqYppT4APjqG/TR5l7a/lNfXvM7k1ZMbPoABcMZtUF4CPz4FwWEw5DEJYgjRjAVVq6dx2SlGK1dnuZv1ewv9AY3VWfnMX2t8rTUpaBcf5g9odE+NJL1FOEFST6PeHHMAQ2udp5TqBgwBbsJI87y1luHjMNIye/nudOwEjqy8p4DHtdavHcPhTwF+01pvB1BKzQQuz3/lKgAAIABJREFUBqoHMC7GuPuiMTI+IpVSiVrrfcd2hkKIE+V0V2AqV6Tfa2ROt9+/lembPsGRs4/iAefyVKdF2K2lzHtuISme3XjCLER6w8iIHU1BxV7WFn3J69ck0zuxFxXeCnpNL+bKvt24sfuZAT4zIZofpVRr4E6gj+8zfhpVn9WzMD7rDwFLtdZFvkzHb7TWl9Wyy5Jqj6cBl2itV/syDQb82XSo4ZpAKXUrtQcD/kjle+YBjyulojGyH77DyG7Ir6WWBRx+HlU71NoFuHyPlyultmFka2QBKdWGpgB7j2Hfbg5fD1f9OqmiWtaqB9/1mtb6el/GyAXAKqVUd6314a2dmhmr2crVGVfz5NInWbZ/Gb1b9G74SQy8B1zFsOQVI4hx9j0NPwchRMDYgy30aRVNn1ZVNd5yi12s2VPgX3qyYNMBZi83kvGCLSa6JEXQLSXSVyjUQauYUEwmCX7WheOpgRELlGutP/R9aE/zLRHJUkpdorX+RCllxVji4QAO+IIXA4Gayvh/BTyilJrhWz+bjPGBfaCGsckYa18rZXF0dkVNY5Ix7phUP4/rMNblkpaWdoxnL4T4M1prSr0eIrAxsWccXT6dTuufv8aSnEzim28SdkY/HNMHUlRaxBcFF+NFc0ZMMuel9uWBQbeRVZTFeR99ye9FO+md2IscZw4aTbw9PtCnJkRzFYHxhbpAKZWAkcn4g++1HzDqXl1LVWbFEuBlpVQ7rfVvvpoSKVrrLTXsOxzY58vIHAfsqbaPEb59jq02vsZrAoyszGlKqScwrlmGATXd+Fjs29903/EWAfj29SvG0pXPfJkVhUqpHUqpUVrr2b7ATFet9eo/+stSSsUBh7TWHqVUG4wMke2+JThFSqlTgV8w6nW8+Ef78tkJ3KiUMmFcr5zyZ29QSrXVWv8C/KKUGgakAs06gAEwqsMopq6byqurX2VKiykNPwGlYOjjUFECPz5ptFY9o7Z7eEKIk0FMmJWB6fH+Zc1aa7LySn1dT4zOJ7OW7mba4p0AhNssdE1xkJkcSbcUB5kpDpIjQ6RI6P/geJaQJANv+T5oAf7p++94jLWyD2NcbIzCqIvxqVJqGbAK2HTkzrTWXyulOgE/+/7hioErgJoCGMeSnnlMKZxa69eB1wF69+4tLc6E+AMer2b573m4Pd6qjdV+0hZtzeGNhdup8Bg/Sh3beunxu+Li927HnZtL9IQJxP3tZkx2o5VqXGIGu/eX4jlzHLvcxZTs/5KuaUYNuqSwJGxmG9sLtgOQ7cwGIMEuS0SEqA++7IiVwHqMZaI/VXvN4ysqeRVwpW/bQV82xfu+GxZg1MSoKYBxH8aX+d+BtRgBDTAyN99VSt2BsRy1wLfvGq8JtNYrlFKzMK4lfgcW1nI6fwem+gqBVhbxrDQLmM3hWSDjgFeVUvdiFBufiVHw84/0Bx5WSrkxsiKu11of8r12A1VtVL/gDzqQVPMTsAPj72cdRn2PP/OUUqo9xm/iBccw52bBZrFxTcY1/Hvpv1m6f2lglpIoBRc+B+VO+PYBo034qdf/+fuEECcFpRSp0XZSo+1c2NWop+H2ePntYLER0MgqYG1WAVMWVV03x4QGk5lidD7pmuyga6qD+PAjFy2II6mm0KZcKXUa8KDWeojv+T8BtNaPVxvzGvCD1vp93/PNwIA/WkLSu3dvvWzZsnqduxCNjder8WiN1uD1/fxrDRqNSanDil6+uXA7kz7f+If7G9IlgfQWEVgLcol++2q6bwVrx44kPvIIIZkZh419aulTvLPhncO2zR42m47RHQEY9ekoIoIjeOj0h/hpz09M+mUSc4bNIT06vS5OXYhmSSm1XGsdgLz64+fL2ijVWmul1FjgMq31xYGeV3PTHK9vytxlnPfRebRxtGHKkABkYVTyVMCcCbDxUzjvKeh7XeDmIoRoclxuD5v2FbEmK581WQWsySpg64Eif5HQFhE2uqY4fH+MIqFRocGBnXSA1HZ901RKpi4F2vvW6+7BSBO9/Igx84CbffUx+mK0F5P6F0JU8/O2XK57ZxlFLnetY05pFU1MmPGLcvG2XPq2jub2wR38r1cPeYYEmclMCqfgww858NTTuJywdnAMo579ABUUdNS+x3UaR5QtisrAaZQtivSoquBE+8j2fLr9U8776DwAFIoWoS1O5JSFEI1LL4w2qwrIB64O8HxEE9EosjAAzEEw8i2YfRV8cZeRmXHKtYGZixCiybFazHRLNQp+VqosEmoENPJZm1XA1xuy/a+nRdvJTHEYS0+SI8lMcRB2Enc+aRJnrrV2K6Vuxlgjawamaq3XV7Zg01pPBuZjtFD9DaON6oTa9idEc7L7kJOC0goAnOUe5q/dh7PcCFCUu72s2JWPy20U1M93VpDosHFd/zYoZaS7KWUECkwKCssq+H7TQfJLywGjVdS9F3QmM8VR47Fd23ew+8q/4Vy2DHuf3vyj+3JOb5NUY/ACjGUiEzMn1nout/S8hb6JVeVtEkITcFhrPrYQounxtR3tFuh5iKZpZIeR/loYAQtgwOFBjPl3GtskiCGE+B/VVCS0oLSC9XsKjKUne/JZtSufz9cY9+aVgrZxYcaykxQHmSmRdEmKOCyLujlrEktI6ktzTLEUzc/uQ04OFrsAoybFsp15lPgyKHKKXcxcuvuw8VaLiWhfqpkCuiQ7iPE9t5gVE/q1pm1c2AnNSVdUkDtlKjmvvIKy2Uj4x104zhtIz9lncWVML24d9vYJ7V8Iceya0hIS0TCa8/XNjI0zeOLXJ5g6ZGpggxgA7nL44C+w5Qu44Fnoc01g5yOEaNYqO5+s9WVqrM4q4GCR8R3BbFJ0SAgnIymCzBQHGckOOic27aBGU19CIsRJY9HWHPbmlwJwoKiMZ7/Z4l8XV8nsa8OkgBE9Uxia0cL/vGfLKH8Aoz6UrlnDvnvvw7VlC+FDh9Linn9hiYvDc2g7bqWwBtnr7dhCCCFObiPaj2DK2im8suoV3hr6VmAnYwmG0W8bQYzPbwdlgt6SACyEqB9Hdj4B2F9QVlVPY0/BYe1czSZF+/gwMpId/sBG50QHIcFNN6gBEsAQosEVlVVwwBctBUiJCsFqMX6R/J5bwhVTfjls/Cmto7lhQFt/84+OLSJo4Wj4CsXekhIOvvACh6a/iyU2lpSXXyL8nHP8r5eX5QNgtUgAQwghRP2wWWxck3kNT/z6RGBrYVSyWGH0OzBrPHx2K3jdspxECNFgWjhstHC04Nwuxs1MrTX7CspYu6eAdXsKWLungB82H2COL6hhUtDOH9Rw+IIaEYQ2oZoaTWemQjQDucUuhjy3kJziqgBG27hQbj67HQrFd5sOYFIw7+Yz/BWHEyNsmEyB7RFdvHAh+x94kIq9e4m8bCzxt9+OOTwcnIdg9UzwuinPM9qfWoNDAzpXIYQQzdvIDiMbTxYGGEGMMdOramK4y+D0vwV6VkKIk5BSiqTIEJIiQxhSLaixv7CMdXsK/YGNhVtz+GjFHt97jJoamckOf7ZGl+TGWyi0cc5KiGZCa81dc9awbk8BYBTkyXeW8/ilmdiDzRSVuXl8/kZum7Xa/56hXVqQkdw4Cle68/LIfvxxCud9SnCbNrSc8S72Xr2qBvz6BvzwGAAusxnSkgkOk64hQggh6o/VbG1cWRhQlYnx4UT4+l6oKIX+vi4lQggRQEopEh0hJDpCGNw5wb89u7CMtVkFrNtrBDUWb8vh45VVQY3WsaFkJjvITHbQJclBRnIE4baaC/U3JAlgCFGPVmcVMGd5Fr1aRhHra016XkYil/RI9o8Z1i2JQyXl/udJkQ2/PORIWmsKP/uM7Mcex1NcTOyNNxDz179isloPH5i1FOI6wsQFuIqy4PNRWB0pgZm0EEKIk8bIDiOZunYqL618iWlDp6EaQ6DAHAQjpkBQCHz/qBHEOOd+CWIIIRqlhAgbCZ1tDKoW1DhQVGYsPckqZN3eAn7dcYi5q/b6X28dG0pGsoPM5Agykhx0SXbgCGnYoIYEMIQ4TjnFLnbklPifh1kt/qKZ5W4vD326nvV7CwEodrmxB5uZNqFPrRFLR0hQg/3gZx1cz+Xzr8CJ0VbVguLZ3v/k9C5j/WPKs/aw/8EHKVm0CFundqRlFmNzT4aXJ1MRbGd1/1twx7UHrbEdWEm3NkNQ1jBcTuPXSbC5/gqICiGEEGBkYVzX9Tom/TKJn/b+xBnJZwR6SgazBS5+BSw2WPQsVDhh6BMSxBBCNAnx4TbO7mjj7I5VQY2cYpex9MSXrbHi9zw+XV0V1GgZYycj2cEFmYmcn5lY73OUAIYQf0Brzadr9rFql1Gg0qs1H63IorDMXet7LCbFRd2SsJiNi5XT28Y2inQrgF07vycPNxdWmIlWQbxjKWPT9q84vctYtMfDoenTOfj8CyilSLjnHqLSXaj5t0PaSAiy8cme//Lwyierdhhj5+lwB0MAl9eo62E1WWs+uBBCCFGHLm1/KW+tf4sXVrxAv6R+jSMLA8Bkggv/Y2RiLHnFyMS48DljuxBCNDGxNXQ/yS12sW5voS9bo4BVu/JpGW2XAIYQgXD3nDV8sHw3ANrXvtQebMbsuzBqGx/GLee0J8hsXIjkOcspqhbQ6JIUQbfUyIad9DEqdRmBmCsHPE6HtkOY/nZXnK4CyjZvZt+991G2di2hZ/Un8YEHCEpKgkXPGW8c9jxYw1j29V+J3fcrTysjKnuPPsDbJb9h2bWAHQU7ALBaJIAhhBCi/gWZg7ix+43cs+gevt31LYNbDg70lKooBUMeM4IYC58BtwsuftnI0BBCiCYuJszKWR3iOKtDnH+b16sb5NjyW1SIagqcFXy8cg+ntYmhd8soAFKi7YzsmRLwTiB1oazMWNpiC4nFpExEuCHti93seHAk5ogIkp55mojzz6+6i1VWAMoMvs4iq4t+p0fLgfQa8CwAY9dN45nlz3Dr97f6jxFji2nYkxJCCHHSuqD1BUxZO4WXVr7E2alnYzaZAz2lKkoZNTCCQuC7SeAuhUvfBIsstRRCND8N9V1JAhjipDd31R4+WGZkXOQ7Kyj3eLl7aMdGm0VxIkrLjQBGSGgcJb/+yqSpHhIOFeO45BLi7/4Hlqiow8bvdO7j4cQEyr8Yj0azp3gPl3e83P/6+M7j6ZfcD482amqEWkJJjUhtuBMSQghxUjObzNzc42Zu/+F25u+Yz7C2wwI9paP1vwuC7PDVv4xMjFFvQ1DgC3YLIURTJAEM0ax5vRqP1ni8Gq3BozV780vZfrAYAJfbyz/mrCEu3EqLCBshQWZG906ha0rjaGNa18rKi7CXacqenMyhj+ZiiVR8NdLCrZMer3H8Sudellot9FRmrGYrA1IGcG6rc/2vm01m2ke1b6jpCyGEEEcZlDaITtGdeHnVywxtNZQgc+OoO3WY024yCnt+fju8PwbGvufPbhRCCHHsJIAhmp23ftrBpM834jnGdVhhVgszrzuVlCh7Pc8s8Oyr9vPsZx6czk+JvvpqJsV+RJS7oNbxzgqj28oLZ7+Aw9o8gzpCCCGaNqUUf+/5d2749gY+2voRYzqOCfSUatbnGmM5ydyb4N2RMO4DsIYHelZCCNGkSABDNDv/3XKQ+HAro3qnYlYKkzLWZJmUwmyCCFsQGckOzL51WgkRNn8b1OaqIvsA2ZMeodM3e9kRDz2nzSIkIwPLzC9w5ueAx11jYbGSilIwg93S/IM7Qgghmq5+Sf3oldCLV1a/wgVtLiAsOCzQU6pZ98vBYoUPr4V3LoZxc8AeHehZCSFEkyEBDNHkZBeW8enqvXh9LUIibEGM7p3qLxyzNbuYU1pHc/vgDoGcZuAUH4ByY4mM1pD/za8cePoZdEUFGwfYebpPOednZABGzYpDJgXOXAhPOGpXTk8ZFrNqnOm4QgghhI9Sirt638XYz8cydd1U/t7z74GeUu0yRhjLSWZPgCnnwviPIDIt0LMSQogmQQIYosl5YcFWZvyy67BtCQ4bA9PjKXG52ZNfymXxJ2khyez1MPkM0F5chWb2LY2k9KAVe9++JD78ELMXjCLIXFWh3R4cxm5lgmkXGBdTZovRqz6pOwBOjwu7at7ZKUIIIZqHLrFduKDNBbyz4R1Gp4+mRWiLQE+pdh0vgPEfw/uXGUGMKz6EhC6BnpUQQjR6pkBPQIjjobXmh80HGdQpgfUPDWH1/ecSYbMw89ddrM0q4NuN2QC0T2hGa0rLS6Aou+qPPqK2R0UZlOZDaT5lS9/g4aho3jnUja1ftSC/0ArDEkmb9hbBLVtS5nERoqoFMBypOK12iO1g3P3ZtwY2zPW/7vRWYFcS5xRCCNE03NLjFgCeX/F8gGdyDFr1g6u/MB5PPQ92/hTY+QghRBMg30xEo7dhbyEvf/8bHq+mwuNlT34pNw1sR6jV+N/3gq5JvP/rLr5an+1/T6cWEYGabp3SpQW89WYvsrXbv+2idpfQ5fznjCclufBcJviKba4uDqXvLw7SDh5kTddwXjrLyXDzATopY3lNqbcCmynEvy97SDSlFhtc9p6xYfKZsHeF8djjxomHULO1/k9UCCGEqAOJYYmM7zyeN9e+ybhO48iIzQj0lP5YQhe45ht491KYPhxGvAGdLw70rIQQotGSAIYIOK01+wvLcHuMzIKQYDOxYVVfmt/6aQffbMimdazRbqxXyyjO7VJVr+Ff53dkUKd4f2JCVGgwaTHNo+jk1k0f85+IEOzKgkWZKfK4KMlZxqTKAbm/QUUJ3swrOfDDAcK/XU1FGEQ//yRjhgzj9RmnkpeXDQe3QEgkpdpNiKlqSYjdYsfpdqK1RikFyT1h/cdGloerEKdS2CWAIYQQogm5JuMaPtr6EU8tfYppQ6cZn2+NWWQqXP0VvDcGPrgSLnga+kwM9KyEEKJRkgCGCLjZy7L4x4dr/M/NJsXLl/dgaEYiWmt+3HqQQZ3jeWVcrxrfH24L4pxORxegbA5+3vElAHMv+pgWka246O1elFU4qwYU7qF4r5V936/EnZ3Dhv5pzBig+HjIMACibVHkmXPg5T4AlLWIx1YtIGEPsuPRHlweFzaLDZJ7wfJp8FAkAKWJ8dKBRAghRJMSFhzGTd1v4pElj/Ddru84p+U5gZ7Sn7NHw1/mwpwJ8PkdULQfBt4DjT34IoQQDUwCGCIg9uSXUu72AjBt8U7axoVyw4B2AExZtIM7PljNM19vwaM12YUuzuoQF8jpNhyvlxenncau8gIA1gQp2phttIhsBYDNbKXMlQeA+9Ah9j09jeIlMeQmFPP1rZl8Ff47/ZP7+3cXFZ5CnldD7/8DoGzHLCIdLf2vVwYnnG6nEcDoMhxKDoLbBUDJ/i9JjEip99MWQggh6tKl7S/lvY3v8ezyZ+mf0r9pdNMKtsOYGfDZrfDjU0YQ48L/QFOYuxBCNBAJYIgG9+2GbCa+s+ywbY9c3IWRvYwvyt1THbz03W+Ue4wAR4/UKIZmJDb4PAPBWbyP181Oom1WIiwhBAOXt73E/3qIJYQyz0EKPppD9pPP4CkqYPYZihVDkokIs9GRjgxvN9w/PsoWTVbxHn8qatnezwmxVtUHsQf5AhgVTqJt0WANhzPvqJrPR4uwW5tHPREhhBAnD4vJwu29b+emBTcxa/Msruh8RaCndGzMFrjoRQhPhB+fhMI9MGoa2ByBnpkQQjQKEsAQDW7NngJMCp4e1Q2TUgRbTAzuXLUEpF18OM+N7RHAGQaOs3g/ADemDGbMoKePej2+KJgBH5vZu+s+Qrp1o6Sfk9mOHbx46p0MSB1w1PhoWzR5ZXn+56XuUmxmm/95aJBRV2TutrnEhcQRZApiSKshhwU2Kh8LIYQQTcmZyWdyauKpvLr6VYa1HYbD2kSCAErB2fcYtTE+uw2mDoXLPzCeCyHESU4CGKLBbT9YTEqUnUt7ytKEI5WUHADAbos8bLt2uzk0/V3Gv7ALL5Bw4+VE3fQvlr1trOsNsYQcuSsAIq2RFFcUU+4pJ9gcTJm7zFgq4pMcloxCMXn1ZP+2Mk8Zl3W8DDCWlkgNDCGEEE2RUoo7e9/JqE9H8cqqV/hn338GekrHp+dfwJEKH/wF3jwHLp8FSSfnDR4hhKhkCvQExMln+8ES2sSFBnoajVKJ8yAAoSHR/m1lmzaxc+xlHPj3v9mfHsWTEyDatgA163JK83YA1BpkiLJFAfizMErdpYcFMDrHdGbh2IV8N+o7vhv1HdG2aDYd2gSAV3spdZdKBoYQQogmKz06ndHpo5m5eSYbcjcEejrHr+1AuOZrMAfDW+fDpvmBnpEQQgSUBDBEg/J6NTtySvwtUcXhnKW5AISGxOJ1uTjwn+fYMXIUFXv3kvzsMyy+ZQDZMSGgvVC0D2eEURuktgyMaJsRCMlz5aG1psxTdtRYh9VBnD2OOHscHaI6sPnQZgDK3GXGXCzybyWEEPVBKTVUKbVZKfWbUur/anh9nFJqje/PYqVUt0DMs6n7e8+/E2mNZNKSSXi8nkBP5/jFd4KJCyAuHWZeDksm//l7hBCimZIlJKLOVXi8fLp6L6UVxkXCnrxSPliWRbnbgwZKKzy0iQsL7CQbKWfpIQDs2/LYcfsllO/ciWP4cOL/cReWqChsv66hNDgErvwRgNKtH8Pi+2vNkqgMYEz4cgJmkxmv9tYa7ABIj0pn5uaZ5JTmUOAyOqFIBoYQQtQ9pZQZeBkYDGQBS5VS87TW1dMEdgBnaa3zlFLnAa8DfRt+tk1bRHAEd/a+k38t+hcfbv2Q0emjAz2l4xeeAFd9Dh9eC1/eDYe2w9DHwWQO9MyEEKJBSQBD1LnvNx3g9g9WH7btrA5x/qwLq8XEBZknR1eR4+XMy2Hilx4sK59DJyeTOuVNwvr1879uM9tw+VqcglGjAmrPwMiIzWBi5kSKy4sBoyr70FZDaz1+x5iOuDwuBn4w0L8tPDj8hM5JCCFEjU4BftNabwdQSs0ELgb8AQyt9eJq45cAUjzqf3Rhmwv55LdPeG7Fc5yddjaxIbGBntLxCw6FMdPh6/tgycuQ/ztc+gbYpFuYEOLkIQEMUec27S8C4Me7BmILMmExm4gODQ7wrBqR/N1QtK/qeWwHCImkaMECkp74idQijXXcKFrd8X+Y7IdnP9gsNsq95Xi8HswmM6XuUqD2LIlgczC39LzlmKc2uOVgXG4XLo8RJLGarTV2NxFCCHHCkoHd1Z5n8cfZFdcAX9T2olLqOuA6gLS0tLqYX7OilOKeU+9hxLwRPLvsWR4787FAT+l/YzLD0McgujV8cTe8OQjGvgex7QI9MyGEaBASwBB1bkt2EanRIaTFyNKDo1SUweR+UFbg3+RuMZD9m1pT9OWXVCRYuX+4mXfuuBNT8NF/f5WZFi6PC7vJjrPCiVmZCTbVTYDIarYyosOIOtmXEEKIP6Rq2KZrHKjUQIwAxhm17Uxr/TrGEhN69+5d435Odm0cbZjQZQJvrH2D4e2H06dFn0BP6X93yrVGTYwProQ3zoYRb0KHcwM9KyGEqHdSxFPUuS3ZRXSIl2UHNdr+vRG8OPdR9Lg55JedzrZXNlK8YAFxt97C0olxbEtStXYVsZqtAP7Mi1J3KSGWEJSq6TpYCCFEI5YFpFZ7ngLsPXKQUqor8CZwsdY6t4Hm1mxd2/VaUsJSeGDxAzgrnIGezolp3R+u+wGi0uC90bDwWdASuxJCNG+SgSFOWE6xi/vnrqO03Cjaue1gCWd3TAjwrBqRr++F3b8ajwuywOagPPF89j38CM6fdxISW0Hi089gPfVCnG/PwKaNWhU1qWyBWuYxOoQ43c5agx1CCCEataVAe6VUa2APMBa4vPoApVQa8BEwXmu9peGn2PyEWEJ4uN/DXP3V1by48kXuPuXuQE/pxES1hKu/hnk3w4KHYP8auPhlo16GEEI0Q5KBIf4nXq/G7fHi9nh5/cftfLluP7kl5eSWlNMtxcF5GS0CPcXGY+lUKNwHQSHo6HYcKBvK9uGXUrpmDY7b/0rqOblYF98Nr5xOSfE+7Kr2iuL+AIavxWlpRSkhQbV3FRFCCNE4aa3dwM3AV8BG4AOt9Xql1PVKqet9w+4HYoBXlFKrlFLLAjTdZqVPiz6MTR/LjI0zWJ69PNDTOXHBdhgxBQY/DBvmwpRzIW9noGclhBD1QjIwxHH7PbeEoc8t9LdJBRjWLYkXL+sRwFk1Up4KqCiBnrdSlnARe++5F9f6hSxrp3hziIlD1ilc2OUMHvdGAlBSYcYeVHtcMcRsBCsqAxiSgSGEEE2X1no+MP+IbZOrPZ4ITGzoeZ0Mbut1Gwv3LOT+n+5nzkVz/rDFeJOgFPS7BRK6wJyr4fWBMGoatDkr0DMTQog6JRkY4rh9syGb0goPfzu7HXcM7sBdQ9K55/xOgZ5W41SaT442seGTX9k+YiRFWTv4zyUmDj5wDRPPuZteCb1Y6C3AO2Y6jJ2BM7kHofa4WndXmYFxZA0MIYQQQhw7e5Cdh05/iF1Fu3hx5YuBnk7daTcIrv0ewuJh+nD4+RWpiyGEaFYkA0McE5fbQ7nbC8CPW3NoExvKHeemB3hWjZ/z50Ws/zqB+Lz1fJ+pmH62i/jEdrzW+3aUUoQFhXH/4vvZUbCDtpFtcVY4a22JCjXUwKhw4rA5GuRchBBCiOakb2JfxqSP4d0N7zK45WB6xDeTTNKYtjDxW/j4evjqn0ZdjAv/A7LkVAjRDEgAQ/yplbvy+MuUXylyuf3b/nJaywDOqPHzFBVx4JlnyJ85C+VQfHZtW04ZfRdPAOnR6f6uIT0TegLw0M8P0cLegs15m+kS26XW/drMR9TAcJeSaEms35MRQgghmqnbet3GwqyF3PfTfcwZNsd/o6DJs4bD6Onw41Pww2NwcBOMmQGO5EDPTAghTogEMMRR8p3ljHltCXnOcgCKytxE2YO4ZVB7AJRSDOsqX5pgx/xpAAAgAElEQVRrU7RgAfsfehh3Tg5RFw9gfLuFXNWhG2elHr0ONS08jYGpA9lR8P/s3Xd4VGXax/HvmZ6ZZGYy6T2EloTepYgCYkFUXLGhrm0V26qLvfvadnVdXXcpdsWyK6vrigVREQUpAqH3llCSkN4mmZlkJnPeP04YREBdN2EScn+u61yTM6fknlAy+eV57qeAal81TrOT0Smjj3nvg9NFiuuLKXQX4va7ZQqJEEII8SvZjDb+b+T/cd2X1zF97XTuHHJnuEtqPTodnHoPJPaGD6+Hl0bD5NelL4YQokOTAKOTUFX1sCmQOp1yzHPnbSxhe6mb8wekYDHq0CkKlw3LIDfZfhwq7QDKt8OBDYf243MgsTeBigpKnngS9/z5mHv2JHXGdNTAeho3LCHiGNM8FEXhb2P/9os/daQpEoA/5/2ZP+f9GQC7Sf5chBBCiF/rpKSTuLDHhby15S1OyziN/vH9w11S68o+G65bCHOugLcnwbiHYeTtWuNPIYToYCTA6ASagypjnv2WfVWe0HO/H9st1MNCVVVmfrubcncjAIt3lJMVa+O5i/qFpjqIH/jnJVCVH9pVo1KozXyC0meeQfV4iLv9NmKuvRbFaKRy6QIAIizRrfKpXRYXL532EuXeckALQEYkj2iVewshhBCd1R2D72Bp0VIeWPIA75/z/k/2o+qQ4npqIcbHv4cFj8L+VXD+LJA+WkKIDkYCjE6g1utnX5WHsdnx9Et18u81hawoqAod31Pp4c9fbMdq0mPU61AUuGN8DwkvjqZyN4V1e9k+9ApIH45+1XySZi8jWPIAEYMGkfT4Y5izskKnez3a1zkionUCDIARKRJYCCGEEK3JZrTxxKgnuOaLa3h+9fM8cNID4S6p9ZkjtSkkaUPhywfh5VO1PhmJvcNdmRBC/GISYHQCNS29LM7tl8ykASnsqWxgRX5l6PieygYA3r52KIMyXGGpsV0rWAz532ofl23j7rhYNpd+y4RPv+HixUE8OiOZt/4W5w33oOgOX5nY29gSYLRM/RBCCCFE+zQkcQiX51zOO1vfYWz6WIYnDw93Sa1PUeCkGyF5ALx/Fbx6Gpz9FxhwWbgrE0KIX0T386eIjq7G6wfAEWEEIM1l5UCdL7Qs6t4KLcBId9nCU2B79+k0+O45WPoC7PwCpcbM39+L5LcLg5TlxHLftTqihyUfCi+a/eCpAk8V3oYKAKyGE2woqhBCCHECum3gbXRxdOGhpQ9R11QX7nLaTvpJMHUxpA6GuTfBh1Oh0R3uqoQQ4mdJgNEJ1HpaAgxrS4ARHYGqQnGNF9CmkNhMemIjTWGrsd1qqITKnXDaIwTvKaYk8gHunR3AUd1EyvPPkX//pRQ59fjXzIYvHoD598PzveCZLvBMF7wF3wDISiFCCCFEB2AxWHhq1FNUeCt4euXT4S6nbUXGw2/nwqn3w8Z/wUunQPG6cFclhBA/SQKMTqC2ZQSG8wcjMAD2V2tNPfdVeUiPsUnPi4NUVRtF0eyH/d8D4KlxUnDeJKpffoXveinsnPF77GedhaOlt0VtxXbIewNWvwmONDjzT3Dm03gHXw1IgCGEEEJ0FL1je3Nd3+v4ePfHfLnny3CX07Z0em2p1Ss/Bb8XXhsPy6ZDsDnclQkhxFFJD4xO4GAPDKdVG2FxMMD4w5x1WE0GSmp9jM2OD1t97c5n0yDvdQCamxTKNkRT894zGFNT4a+PMKv0Sf4anw6Aw6R176678TtinVlH3Mpb8DkUzZMAQwghhOhAru97PcuKlvHoskfJickhLSot3CW1rcyRcONSmHsLfPkAbPsUzpsBMV3DXZkQQhxGRmB0Agd7YNgtWl6V7LBw85iunNw9jkEZ0ZzdN4mrR2aGscJ2pjAPYnvidv2W/IVdqcm34rrqKrI+nkt572QAYiJiAHCYtQCjtqn2qLfyBrRpOhJgCCGEEB2HUWfkmVOeAQXuXnQ3/mZ/uEtqe1YXXPIuTHoRSrfArJGw4iUIBsNdmRBChMgIjE6gxuMnymLAoNfyKkVRuOuM7DBX1X4Fysoo2ZiIe+MCzD17kvrK40T06QNAZZG2ekucNQ74QYDRKAGGEEIIcSJJiUzh8ZGPc/s3t/Pc6ue4Z+g94S6p7SkK9L8Usk6Bj2+Fz++GLR/DedPB1SXc1QkhhIzA6AxqvX6cLQ08xbGpqkrN+/9i9/tQv7WCuNtvp8sH74fCC4AKr7aqSIzlRyMwJMAQQgghTjjj0sdxWc5lvLP1HRbuWxjuco4fezJc9j6cOx1KNmijMVa9KqMxhBBhJyMwThDvrtjL1gOHlvsal53AmJa+FjWeJpwRssLIT2nML6DkkUfwrFqFNc5P4p3XYD5vKgAPLX2IubvmAqCiEmWKwmKwAD8fYHj8HhQUzHrzcXgVQgghhGht0wZNY23ZWh5c+iAfuD4gOTI53CUdH4oCA6+ArFPh49/DZ3fAlrlaqBGdEe7qhBCdlAQYJ4BgUOXRjzdj0OmwmvQ0BoL8c+V+zuiVgKIorC+spVeyPdxlth+NbijbCoDqD1Dxn2+pfPM9lIgIku66Ace+h1G6a1NsVFVl4b6F9Intw/Dk4QDkxOSEbhVpjESn6H6yB0aEIUJWeBFCCCE6KJPexLOjn+WiTy/irsV38eaZb2LUdaKRrc40uOI/2kprXz4Is0bA6U/AoKu0kEMIIY4jCTBOALVeP/5mlXvP6sm1o7rg9vn5w5x1bC9xAxBtNXJaTkKYq2xH5t4MW+biKTdxYJWDpjoj9jNOI+GhRzDUrIN3gEjt61VUX0RdUx2Tuk/iwh4XHnErnaLDbrJT0lBCmacMgGhLdOiNzcEAQwghhBAdV5o9jUdHPMqdi+7k72v+zrTB08Jd0vGlKDD4aug6Fj6+BT69XRuNcd50cKSGuzohRCciAcYJoKK+EYDYSG2aSJTFyKtXDglnSe2Xr47mjV9Qtqc/NavKMMbaSRtdQOS1J0NsLBSWoQJusxWa6lhTtgaA3JjcY97SZXHx8e6P+Xj3xwDkuHJ4d8K7GPVGCTCEEEKIE8QZmWewqmQVb2x+g8GJgxmdOjrcJR1/0RlwxVzIew2+ehhmDocJf4Z+l4S7MiFEJyEBRjv1fX4la/fVhPaHZEYzONN11HMr6psAiIuUPgtHteZtWPoCqhrEvcNH6RIHgaZyXFddRdzNN6F7aRB8+0fY9hlU7ubhWBcfzb8sdLlJZ6K7s/sxb//UyU+xpXILABWeCmaun8lFn16EzWhjT90e4q3xbf4ShRBCCNH27hpyF+vK1vHAkgd4/5z3SbQlhruk40+ng6HXQbfT4KOb4D9TIf9bLcgwR4W7OiHECU4CjDD5Pr+S73aWh/Z7Jzs4q09SaP++DzdSUNEQ2jcZdMyYMhBHhDY1oUusjbgoLbAIjcCIkgADgOo92pJfqBBshkXP4DdlUrJMR/02L+ZUF44//QVj71xqCWAdfjPmte+G+mJstsfSw5nOpG6TAMhyZGHSH7sJaq+YXvSK6RXa1+v05JXkAZDrymVM+pg2e6lCCCGEOH7MejPPnvIsF396MfcsvofXzngNg66Tvp12dYErP4HFf4bFz8DepXDOC9o0EyGEaCOd9H/c8Htq3lY2FNZi0Ck0qyqRZsNhAUZpnY+rRmRy71nZVDY0MeGF77jurbzQ8b6pDj6+ZRTwwykknSjAaA4c+ljRab8NOGjZ37WlvgA1CNX7Eylf34SqQvzdd/NSz0L+ufkG2Kyd7rK4WHDT0lDfirL3RnFm/ACuyL3iV5V2fd/rub7v9b/qWiGEEEK0b5mOTB4e/jD3fncvL6x5gTsG3xHuksJHb4Ax92mhxdyb4e3zYcDlcPqTEOEMd3VCiBOQ7udPEW2huMbHxYPT2PXUBO46oyduX4CGRu2H8obGAJ6mZhLsFixGPSnOCL76w2je/d0w3v3dMC4anMqmolrqW86vqG9Er1NwRnSSjtjr/gmPx8LjMdr2dAbUHxrNQkMFxHTHN3kRe7aMp3S5joiBg8j65GNcV1/FNwcW0zeuL/cNvY/zup5Hla+K4vpiAHwBH7WNtTLtQwghhBDHdHbW2VzS8xLe3Pwm8wvmh7uc8EsfBjcsgVHTtPdpM4ZpU3OFEKKVSYARBv7mIJUNjSQ6LAAkRGmPZW5tJMXBERVxP5gSEm+3MLJbLCO7xTKhTxJBFdbv13pkVNY34bKZ0Ok6yVJWJRvBYIYxD0L/y6GxDqp2hw4H6yopy9NTcMkV+IsPkPzss6S98jKm1FSK6osoaShhYtZEpuRM4YIeFwCwt24vAOVeLQiRAEMIIYQQP+XuIXfTP64/Dy97mG1V28JdTvgZLXDaI3DdQrDFwXtT4P2rtV8sCSFEK5EAIwzK3I2oKqEA4+BjaZ0POHJVkR8bkB4NwOOfbuF3s/NYuK2sc00f8dWANRZOuUtrIgWhb44Ny5aR/0oBlSvqcJx3Ll3nfcaXPTyc/u/TGf/BeKZ8NgWAIQnaKi3pUekA7HfvBwgthSoBhhBCCCF+ilFv5LlTn8NusnPTgps4UH8g3CW1D8n94fpvYOyDsO1TmD4ENrwPqhruyoQQJwAJMMKgpFYLKkIjMOxa+HAwwCh3t6wqcoymnI4II5cNS0enKBTXeImNNDOpf3Jbl91+eGsOzau0xQIQOLCP4nvuYd8116IQJP13fUl+8kn0Tidf7/2apuYmTko6iVPSTuGGfjfQ1dkV0Ppf2Iy20AiMgwFGgjXh+L8uIYQQQnQocdY4Zp42E2/Ay01f34S7yR3uktoHvRFG3wVTvwNXFnz4O3jnN1CxM9yVCSE6OGniGQahAMOuBRjx9sNHYJQfnELyE6Mqnjy/T1uW2K7V+6pYZjESLJgPfh+x+xxE3fkyQX+QmBumEut+Fl3vrND5e+v2MjRxKI+PfPyIeymKQnpUOsuKl/Fc3nNsqdKWQ5URGEIIIYT4JXpE9+D5Mc9z41c38odv/8CscbMw6jtJX7KfE58N136pNVdf+ATMHA7Db9bCDXNkuKsTQnRAEmC0ksc/3cLCbWWhfUeEkWFZLvSK1pciPsrM+F6JGHQKu8vrgUMBRpTZgNWkp7SupQeGuxFFAZft2Et3dmb/CFTwd1Mj6R/cyXXzm7EV2SDLStbf3sLcJQMefwQitGk2/mY/xQ3FTMiacMz7jUoZxdtb3uaf2/4JQO+Y3kQa5ZuqEEIIIX6Zk5JO4tERj/Lg0gd5dPmjPDHyCRSlk/Qm+zk6PQybCr3Oh68egaV/hY3vwxlPQu4kkK+TEOK/IAFGK0l3WemT4gjt7y6v5/UlBaF9f7PKo59sCe3bTHqcVi2dVxSFBLuFD9cUsmpPFcU1XlxWEwa9zPA5mnqvl6uWwIQ8CNiszDjbx++GpWPu1u3QaiQtAUZhfSFBNUiGPeOY97t14K3cOvDW41G6EEIIIU5Q53U7j+KGYmaum0lyZDI397853CW1L5HxcP4sGHQVzLsD3r8KupwCZ/4REnqFuzohRAchAUYruXJEJleOyDzm8fX7a9hcXBfa7xpnOyyZv2ZUFxZuLQUgxmZiWFZMm9Xakbm//ppT3mrGWq/gvPBCSn87nkXLbuLSRm1FFrzVlOj1fO3eAVvfpaBWC5F+KsAQQgghhGgNN/S9geL6Yl5c/yLJtmTO735+uEtqf9KHwfWLIO91WPg4zBoJ/S+DMfeDIyXc1Qkh2jkJMI6TfmlO+qU5j3n8ipMyuOIk+SH7WPxFRZQ8+RT1CxfS6FJ59cpIZt33GO7qHQDUN9ZqJ3qrec1p573C+VCorcseaYwky5F1rFsLIYQQQrQKRVF4ePjDlDaU8tjyx3CYHYxNHxvustofnV5bSa73BfDdX2Dly7Dp3zD8Jhj1BzBHhbtCIUQ7JXMURLum+v1UvvYauyeeQ8Py5cTfdhP/vjTAgUwbQKhXhcdXC2VboXQjNTod6REJLLlkCUsuWcK3F39LlEm+EQohhBCi7Rl12vKquTG53LHoDhYXLg53Se2X1aX1wrglD3LO0cKM6UNh80ey7KoQ4qgkwBDtlmfNGgp+cwFlf34W2/DhdP30E2IumkCDQcGi11ZosRm1IKOhqQ5mngSf3UGdTofT4sRhduAwOzDrj72aixBCCCFEa4s0RTJr/Cy6O7tz+ze3s6pkVbhLat+iM+CCV+DaBWCLgfevhDfOgn0rwl2ZEKKdkQBDtDuB6moOPPQQe6dcRnN9PakzppM2cwbGlBTw1eBTFCIMEQBYjVYAGvpOhgvfhAvfxJ2QQ1REbBhfgRBCCCE6O7vJzsvjXyYlMoXbvrmN3TW7w11S+5c2BK77Fs5+Dqry4fXT4Z9ToGxbuCsTQrQT0gNDHH+eKljxEjRry8ZisMBJN6Gao6idO5eyp5+hua4O1zXXEHfzTegOfA8v9Ae1GQKN+OwK0S3BhVFnxKw30+BM15bnAup2vEaqyR6uVyeEEEIIAYDT4mTWabO4fN7l3LjgRmafOZukyKRwl9W+6Q0w5Frodwl8PxOW/g1mDYd+U2DMfeBIDXeFQogwkgBDHH/bPoNFfwKdUVv7u7mJQJOJAx/tpn7hQiL69yfx/x7F0rPnofPrSyH3PAC8DetItsWHbmcz2mjwN4T23U1u6XkhhBBCiHYhNSqVGafN4LovruPqL67m9TNeJzkyOdxltX8mG4y+CwZfe6jR58b3Ydj1MGqa1j9DCNHpyBQScfy5D2iP9xfBQ+W467qQf+/rNCxZQvy995Dxj3cPhRcAZVsgqR+c/yKc/yI+i4MIU2TosNVgpSGgBRiqqlLXWIfdLCMwhBBCCNE+9Irpxcunv0xdUx3XfHENRfVF4S6p4zjY6PP3q7VVS5ZN10bmfvccNHnCXZ0Q4jiTAEMcf+4SiHDR7AtQ/OCDFM5rxGDykXmhmRjPSyivjoWKndq5qgqlWyA+J3S5r9mHRW8J7UeaImlo0gIMb8BLQA3ICAwhhBBCtCu9Y3vzyumv4G5yc/X8q9nv3h/ukjoWZzqcPwtuXAoZw+Hr/4O/D4S8N6A5EO7qhBDHiQQYovXVHYB3LoA3Jmjb7HO1EOIgdwkedywFkyZR++F/iLniQrrcNhJLjx4Q01ULLz6/G7Z+Auv+AY21EJ8butwb8IaaeMLhIzDqmuoArXGWEEIIIUR70iumF6+e/iqegIer51/Nnto94S6p40noBVPmwNXztVDj09u1lei2zJWlV4XoBCTAEK1v90LYtQCCAVB0ULAYtn4MgNrURNnnO9j7QT0oChnvvE38A4+hXDobLn5H20bert1jzuUw9ybtnimDtOtVFW/Ai8VwaATGD3tgHAwwZASGEEIIIdqjnJgcXj39VfxBP1fOv5KtlVvDXVLHlDEcrvkCLvkH6PTwr9/Cq+O0951CiBOWBBii9VVsB70JrpoHV30KMd2gZCO+HTsouOhiKlc2EBgUQ/mse1mf2Mia0jUE1eCh60+eBjcuhxuWaNutayFlIACNLSuXHCvAcDe5ARmBIYQQQoj2K9uVzewzZ2PSm7j2i2vZXLE53CV1TIoC2WfDjcvgvBnaNOXZ58Db58Pub2REhhAnIAkwROsr3w4x3bVlsAA1vjeVX6xjzwWTCZSXUzHew5TTa7l+6W1c/9X1XDn/Sr7Z/82h63V6SMiFxD7a5soKHfIFfACHTSGxGW24m9zsd+8PDcWUAEMIIYQQ7VmmI5O3znwLu9nO9V9dz7aqbeEuqePS6WHA5Vqjz/GPw4EN8PYkbWpJ3uvQ1PDz9xBCdAgSYIhWt79yKxdEBjj7w7O5/LXTmff6asqWB7HlxJM17ST2dGkGYNZps3jjjDcw6AxsLN/4i+7tDXiBwwMMp9lJla+KCR9O4NHljwIQbYlu3RclhBBCCNHKkiKTePX0V7Earfzuy9+xaP+icJfUsRkjYOSt8IfNMGmWNiL40z/Ac7mw8AloqAh3hUKI/5Eh3AWIE0BtkTZUryXdXh2sZkfQxY1bIxj5/j4CwQAFp/o5K3ENyqY11Lii0aEwInkEOkVHd2d3tlb9svmf3uYjA4yrel1FV2dXVLRhgtHmaFlfXQghhBAdQmpUKq+f8TrTvp3GLQtv4bKcy7hj8B0YdcZwl9ZxGS3Qfwr0uxT2fQ/Lp8PiP2tLsA64HEbcAtGZ4a5SCPErSIAh/ndlW7S+Fz3OAmsMZZX53P3vvQzeuZmIIUO4buh6Joy4ngmD/gBA7fdPYN/zBTpFGwCU7crmm/3fhEZhmPQmekT3QFGUIz7VwSkkP1xG1Wlxck7Xc9r6VQohhBBCtIm0qDTenfAuz69+nne2vsOeuj385ZS/YDPawl1ax6YoWrPPjOFQvgOWvQCr39SmlfT+jdY4PrF3uKsUQvwXZAqJ+N81ait/cNqj1NkmMfjFYvoVqMTfew8Zs98kmBhLla8qdHpNYw1OszO03yeuDzWNNUyZN4Up86Yw+ZPJfLn3y6N+qoNTSH7YxFMIIYQQoqMz6U3cM/QeHhn+CN8Xf89V86+i0F0Y7rJOHHE9tEaft2+Ak26E7Z/DiyPhn5dC4epwVyeE+IVkBIb43/nqaG5SKH16FrWffUlNagTzbxnAM1deBUCMJeaIAMNhdoT2J3WdRGpkKv6gH4Bn857lr6v/yrqydYAWWhTWF4IKtU21wOFTSIQQQgghThSTe0wm0ZbI3Yvu5uJPL+aPJ/+R0amjw13WicOeDGc8CaPvhBUvw/czYftYSBkEQ6+H3EnaFBQhRLskIzDE/6x+zRby58dRO/9rHDdcx2NX27B27xE67rK4qPIeCjBqG2uJNh9qsmnUGxmePJzRqaMZnTqa2wbchtvvZu6uuczdNZeF+xbiDXjxB/1YDVZGp46mm7PbcX2NQgghhBDHy6iUUcyZOIckWxI3f30z09dOpznYHO6yTiwR0XDqPXD7RjjrGfDVwX+mwvO5sOBRqNkX7gqFEEchIzDEr9ZcV0fp009T++9PKXYZmD4ZdkW/AQHIsGeEznNZXBTUFoT2q33V9Izuecz7jssYx7iMcW1auxBCCCFEe5ZmT+OdCe/wxPdP8NKGl9hQvoGnRz8tK621Nosdhk3VRl8ULIKVr8DSF7Stx5kw5HeQNQZ08ntfIdoDCTDEr+JeuJCSRx4lUFWF99Rk7hpSygW9LuWcyCQMioFzu54bOtdlcVHlq0JVVRRFobax9rAeGEIIIURnpCjKmcALgB54VVXVP/3oeDbwBjAQeEBV1WePf5UinCwGC4+PfJz+8f3544o/ctGnF/GXU/5C37i+4S7txKMokHWqttXsh9VvwOrZsH0euLpqQUb/KRAh72GFCCeJEoXGXQJlW1u2bdAcOOppgepqiu64k8KbbkbvcpE5Zw5VY134DQqTe0zmmt7X8Ntev8VpOfSfuyvCha/ZhzfgxRfw4Wv2HXZcCCGE6GwURdEDM4CzgFzgUkVRcn90WhVwKyDBRSemKNp7rLcmvIVe0XPl/Ct5b9t7qKoa7tJOXM40GPcwTNsCv3kFrDHwxX3wXA58fCuUbAx3hUJ0WjICQ4C3Gv7aB5qbDj03ahqc9khoV1VV3PPnU/L4EzS73cT+/hZir7sOxWTCu74BOHZjTZfFBcBdi+9C15KZ/bCJpxBCCNEJDQV2qaqaD6AoynvAecCWgyeoqloGlCmKcnZ4ShTtSa+YXsyZOIf7vruPJ1c8yZqyNTww7AF5T9WWDGboe5G2HVivTS/Z8C9YMxvSh2ujMnLOBYMp3JUK0WnICAwBdcVaeDH8FrjwTXBmQPn20GF/aRlFt95K0R+mYUxOpssHHxB3880oJu0/a6//pwOMAfEDyI3JpchdxH73fnJcOQyIG9DmL0sIIYRox1KA/T/YL2x5TohjcpgdTB83nVv638KXe77kvI/O46u9X4W7rM4hqR+cNx3u2AqnP6mNXv73tfB8L1j4pPZ+WgjR5mQEhoCGCu2xxxnQZTSs+yfU7kdtbqb6n+9R/vzzqH4/8Xfegeuqq1AMh/+18Qa8YNTmaR5Nhj2DORPntPWrEEIIIToS5SjP/eo5AYqiXA9cD5Cenv5rbyM6AJ2iY2q/qZySdgoPL32Yad9OY3zGeO4fdj+xEbHhLu/EFxENI26Bk26C3Qth1Suw+M/w3V8g+2wYeh1knqz11BBCtDoJMAR4WgIMa8s3PUcqvvWrOHDJpfg2bsQ2ciSJjzyM6RhviLzNjWA89ggMIYQQQhyhEEj7wX4q8Kt/hauq6svAywCDBw+W5gidQLYrm3fPfpfZm2cza90sVpas5L6h9zGhywQU+eG57el00P00baveA6teg7Vvw9aPIboL5J6rTS9JHigrmAjRiiTA6CxqC6G+7NB+XDaYrNrHnircikKDEkQtL8D7eQGNC83oo4tIfvZZ7Gf/9DdCb3MjRkwYdPLXSQghhPiFVgHdFUXpAhQBlwBTwluS6GiMOiO/6/M7xqaP5aElD3Hvd/fy1d6vePCkB2U0xvEUnQmnPw5j7ofN/4GNH8DyGdpSrFHJkHOOtmWMAJ0+3NUK0aHJT5ydgd8H04dCS68KAPpfDpNmAOB2FzMuPYXcv13MtV8Gia2D4twApz77N/RZg6B6L7x7IdSXaNcqejj5Dq1xEeAN+rEo1uP9qoQQQogOS1XVgKIotwBfoC2j+rqqqpsVRbmh5fiLiqIkAnmAHQgqinI7kKuqal3YChftUpYji7fOeovZW2YzY+0Mzp97Pg8Pf5jxGePDXVrnYozQllrtPwW8NbDjC21ExprZsPIlcKTDsOthwOXaVBQhxH9N6cxLMA0ePFjNy8sLdxltr2wbzBwGI2/XOiZ/9yz4vXDjUgB2vnklSz9cxbAdKr6MBP421kN8ZAVPp58DrizYtQD2r4SBV2rz+Uo3w57vQrd/ONbFUmc8X1++IlyvUAghOi1FUVarqjo43HWI9qPTvL8RR5Vfk88DSx5gU+Umpvadyk39b0KnyLnfNFsAACAASURBVBSGsGpq0MKMVa/B3iWgN2m95/pcBN1PB+PR+8gJ0Zkd6/2NjMDoDKp2a4+550LKIO0/zhUvE/Q0UDX7LfwzVtEfFd91k+l/68P4Pp9C7YE6WP3moXuc9iiM+oP2caAR1r2rJcuAt2wRVpoQQgghhBDhleXURmM89v1jvLThJdaUreHinhczJm0MJr0s9xkWJhv0/o22HVgP69/Tppls/QTMDu09et+LIGOU9MsQ4mdIgNEZVOVrj64sANTYbOr3QenEifiLS/BmW7hjrJ+XLr8cxWjEaY2lKrkvXL265QbKoX4ZoK2JPfia0K7363wiPD/oryGEEEIIIcLGqDfy2IjHyHHl8Pqm17lz0Z0k25J5ZPgjjEgZEe7yOrekfto2/nEoWAQb39f6Zqx9G2zxkD0Bss/RVgY0SOAkxI9JxHeiCDRqy6Ee3ILNh45V7oYIF0RE01hQwP6/fkrhdzEozfWk3zOZ3eObqHQoRFu0uXjR5mhqm+q0tNhkOzy8OApfwHfMJVSFEEIIIcTxpygKU3Km8MUFXzBz3ExMehNTF0zl7kV3k1+bH+7yhN4A3cbB+S/CnTth8utak88N78O7F8Cfu8IH12rhht8b7mqFaDdkBMaJYtZIqNx5aN+Zof0nCFCwmGBkJhV/+QuVb85GZzIRP6QRV5dilL07qHHYASdOsxMAh9lBta/6F39qb8BLpCmyFV+MEEIIIYRoDXqdnpNTT2Zo0lBe2fAKb215i/l75nNWl7OY2m8qWY6scJcoTFbofYG2+X2Q/602vWT7PNj0gdbwc+BvtSb8cT3CXa0QYSUBxokgGITKXVoToO6nQzAAmz6EvUtRVZXanQrla90Eal7FMWkS8XdMw+CwaQ2FgOr1M7DtnR+aFxlticYT8NDU3PSL5kp6Ah7irHFt+hKFEEIIIcSvZ9abuWXALUzJmcKbm9/kvW3v8XnB54xLH8fkHpMZnjxcmn22B0YL9DxT25oDWuP8vNdh2XRtWda4HOg6FtJP0qaZRDjDXbEQx5UEGCeCJjegav+JDb2OoBrks7hUvCtW0uXtxUQWlBHMSaTLzMexDhxw6LqWqSE1wcbQ6Asg9HFNYw3x1njKPeVc+MmF1DVpq7bpFB2PDH+Ec7qeA8gUEiGEEEKIjsJlcTFt0DSu7nU1b215i3/v+DcL9i0gPSqdq3pfxcSsiUQYIsJdpgBtmknXMdpWd0BbknXrJ5D3Gnw/A3QGyBgJ2WdDzwngTAt3xUK0OQkwTgS+Ou6Mi2Hz3jlwYB7OUg9nfV7OkJ0q5XZ47VwdK3rtJ2XvQ7BXu+TS7Eu5LOcyAKobq48aYFT7qom3xrO0eCmVvkou7nkxkcZI/rPrP3y7/9tQgOENeOUbnRBCCCFEBxJtiea2gbdxY78b+Xrf17y5+U0eW/4YT33/FNmubMZljOPinhcTZYoKd6kCwJ4Ew6ZqW6AJitfAjvmw7TP4/G5tS+wDPc/WRmikDAS9MdxVC9HqJMA4ETTW8bXNSq7PwAXf6ei5qBzVZMJ1+1SyrricfoYgr258lVJPKQDbq7Yza/0sLuh+ARaDhRpfDU7LoQDjYDPP2sZaAPJK8nCYHdw/7H50io797v1srtwcOl8CDCGEEEKIjsmkN3FWl7M4M/NM8krzWFa8jLySPF5Y8wKvb3ydS7Iv4YrcK0LvD0U7YDBpU0jST4LTHoWKXbD9M9g2DxY9DYv+BKZI6HEm9JkMWaeCUd6rixODBBgngMbqEs5aqXLp8koMjeU4L7qIuN//HkNMDABm4I7Bd4TOX1Wyimu+uIZR741Cp+jwBXxMzJoYOu4wOwB4asVTOC1OtlVtY1jisNC8yF6xvfhy75fk1+QTaYrE1+zDopcpJEIIIYQQHZWiKAxJHMKQxCEAbK7czKsbXuWVja/wztZ3GJs+luFJwzk59WRcFleYqxWHie0GsbfByNugoRL2LoFdX2tTTjZ9AHozZAyHrJbpKAl9QCf9TkTHJAFGB6YGAtR8+CFlf/0LV1QFqe2TxICnZmHu3v0nrxucMJi7Bt9Fmacs9NzErocCjEx7JqdnnE51o7YSSe+Y3lySfUnoeJ/YPgCcN/e80HN2k71VXpMQQgghhAi/XjG9eH7M8+yu2c3szbNZVLiIz/I/Q6foGJo4lDMyz2Bc+jgZmdHe2GIg9zxtm/AsFCyG3Qsh/xtY8Ii2WWMh6xTIPBkyR0FMN1CUcFcuxC+iqKoa7hrCZvDgwWpeXl64y/hl6g5A9R4A1GAQ97oiyl98naa9e9F3S+KhoaVcPPEWfjPwxjYvJagGmV8wn3p/PQAGnYFx6eNCIzeEEEIcP4qirFZVdXC46xDtR4d6fyM6jKAaZFvVNhbsXcCXe79kb91e9IqeiVkTuX3Q7cRGxIa7RPFz6g5oS7Tmf6M91mvTy4lM0JqBZo7UQo3YHhJoiLA71vsbGYHRUbxxJmrVHhpKzJRtiKKx2oS5ezdSZ0yn1LSKLfnvYo04Pgm4TtExIWvCcflcQgghhBAi/HSKjtyYXHJjcvn9gN+zvXo7c3fN5b3t77Fg3wJOST2F3JhcxqSNId2eHu5yxdHYk6D/pdqmqlC5W1umde9S2LMENn+onWeLawk0RmmPcdky5US0GxJgdAR1xXh2FlFe0AfP7kqMThPJo+qwv/gfFIOBPV8tAMAaERPmQoUQQgghxIlOURSyXdlkD83mkuxLmLFuBqtLVzOvYB7P5j3LsKRhjEkbw4D4AfSM7olepw93yeLHFKWld0Y3GHy1FmhU5R8KM/YshS0faedaY7QgI+tUbYUTV5dwVi46OQkw2jnP6tVUPP0oDRvi0EcHSHjoQaIzqlEWPACNNWCIxduyWkjED5ZCFUIIIYQQoq1l2DN4ZvQzAJQ0lDB311w+2vURf1r5JwBcFhejUkaRZEsi0hjJkMQh5MbkosgUhfZFUSCmq7YN/K0WaNTsPRRmFCzWmoICRGe2NAQdC11GQ4T8DCKOHwkw2inPhzMof/NfeHaUobfqiB9QT/SLy9A5omHHF9pJVflgi8XTVAeA1WgNY8VCCCGEEKIzS7QlMrXfVKb2m8qB+gPklebxXdF3fFf4Xag5PEBaVBpnZp7J6Zmnk2BNwKw3y/vY9kZRtKAiOhMGXN4y5WQX7P5G66Gx8X1Y/QYoOkgZpAUZKYMgoRdEJoJRVigUbUMCjHamYeVKKmbMxLNiBXpzM/EDfUR396HLPQMcLT0uXFnaY1U+pA3F06Q104yQ9Z2FEEIIIUQ7kBSZxDmR53BO13MAUFWVmsYavt3/LZ8XfM5rm17jlY2vhM5Pi0qjf1x/+sf3p29cXzLsGUQY5L1tu6EoENtd24ZdD81+KMzTwozd38CSv4LafOj8yEToeZY2SsOeovXfiEqS5qDifyYBRjugqiqeFSupmDkTz8qV6GNcxA+oJXrqPehOve3IC5wZWtq54kXYswRP1S6wKVgNklwLIYQQQoj2R1EUoi3RnN/9fM7vfj4V3gqWFS/D3eSmwd/AlsotLCtexif5n4SucVlcuCwuYiwx5Mbm0i+uH+lR6aREpsiIjXDTGyFjuLaNuR+aPFC6Ccq3QX0ZlGyEDf/SRmkcZLSCqyvEZEF0F3Ckar+YTR0MFlnNUPwyEmCEkdrcjHvB11S++iq+jRsxxMWRcP99OAfFonv/UsgYePQLDSZtbed9K8BdiseiAHr5j1wIIYQQQnQIsRGxnNv13MOeU1WVwvpCNlVsYr97P8X1xdQ01lDmKePtLW/zRlD7YVhBIScmh+FJwxmePJwB8QMw6U3heBniIJMV0oZq20F+L5RtgfpyqCuEynxtGkrJJtg2D4L+lhMVbepJ2jDt+uQBENMNpPmrOAoJMMIg2NRE7dy5VL32Ok179mBMTyfxnltxDEpGZzbB9nnaiYm9Q9e8tP4l1pavDe2f0ucMLr3wTQA861+EdTNkmJ0QQgghhOiwFEUhLSqNtKi0I475Aj62V2/nQP0BCmoL+P7A98zePJvXNr2GRW9hUMIgBiUMItoSTaQpkihjFCoq+TX5VDdWk+XIIsuRhd1kx2FxEGWMkkaibc0YofXFOJpgEBrKtYBj/wrYtxw2zIG817Tjig6ssVoPjuQB2pYyUIINIQHG8dRcX0/NnDlUvTmbQHk5li6JpNz6G6IGZ6EsfAw+rDl0ckx3iNB6XpQ0lDBj3QxSIlOItkRTVF9Efk0+l2ZfCoAn4MGoM2LUGcPxsoQQQgghhGhTFoOFfnH96BfXD4Ab+99Ig7+BVSWrWF68nOUHlrN07dKjXqtTdATV4GHPRRojSY5MJjkymQRrAk6zk3hrPH3j+pJkS6KkoQSALEcWRv3h77FVVaXQXUhADZBpz5Qg5NfQ6SAqQdu6jtGeaw5AxQ4oXgvVBdpUlIqdsPYdWPmSdo4pEpL6a0u5WhwQldjSbLSL9miODMvLUVWVUk8pcRFxnW7ZYFVVWVe+jihjFN2iu7X555MA4zjwHzhA9bvvUj3nXwTdbqwnDSNpaCW2yDUoZWtgHhCZAFd+AuYoAL5w72b1iqcAKKgtQEXl5fEvk2ZP46X1LzF93XQ8fg9WoxWv3yvTR4QQQgghRKdiM9o4Ne1UTk07FYAGfwPuJneor0az2kymPRO72c7e2r3sc++j3l9Pta+aAw0HKKovori+mLVla6lrrENFPeJzGHVGekT3IDcmF7vJTlF9ERsrNlJUXwRAamQq8dZ49tTtwWVxMbnHZIYkDkFVVXbV7GJlyUq2VW1jT+0ezHozPaJ7EGeNw6w347K4SIlMITUqleTIZKp8VWyv2s62qm3srN5JlCmK7tHdGZI4hKGJQzHqjFQ3VuNv9qPX6XFZXOgU3f/0NQyqQXZW72Rr1VacZieJtkQSrYk4zI5QMNPU3ESpp5QkWxIG3S/78TGoBkO1bSjfwNqytVj0FrwBL9urt7O3bi9lnjKcZidT+03ltPTT8Kl+mqPTscXnoKJS4a2gpKGEA/VFRDRU0b+xEXvJFi3g2LUAvDUQ8Ib+1BQAR5o2UiM+Vws04nO0j/Wt+4teVdXqW1mykmXFy1hWvIwKbwV2k51hScMYljiM7Jhs1petZ335eqJMUSRYExiePJy+cX1DodrO6p1sqdyCJ+AhEAwQaYwkJiKGfnH9iLZE/081evwe8krzqGmsIduVTZYj67A/v9rGWnbV7GJ3zW521+wmvzafgtoC7GY749PHk+nIxN3kpt5fT31TPRn2DMZnjMdqtKKqKkX1RawqWcX7O95nY8VGzsk6h6dOfup//dL+LEVVj/yH2lkMHjxYzcvLa5N7q6qKd+06qt56C/dXX4EaJKpPIjFjuxLh8sPm/8B5MyFzlHaBLU6bO9Zy7eg5o/EGvFgM2hJEI5JG8Mwp2hrbX+39imnfTmPOxDnkxuTywJIHWFWyii8nf9kmr0UIIUT7pSjKalVVB4e7DtF+tOX7GyFOVEE1SFF9ERvKN1DprSTBlkBQDbK1citbKrewpXIL3oCXpMgkujm7MSJ5BDpFx+LCxdQ11ZFpz2Rn9U42VW467L5Rpij6xPYh056Jr9nHjqod1DTW0NjcSJWviuYfrtzxg2u6O7tT768nvzafQDBAhCECVVXxNftC50UYIsh2ZZPjyiHblR0KRrZUbmF71Xb6x/fnjMwzUBSFCk8Fbr8W7Hj9Xip9leSV5LGiZAVVvqojarDoLSTYErAarOyq2YU/6MekM5HlzKJHdA8y7BkYdUZ0ig6doiMQDFDprWS/ez+bKjZR21RLv7h+NKvNrC5dfdi9E6wJZDmyiLPGsaliE/m1+Zj1ZhqbGwFCP2QHgoHDrlNQSLIlEW2JRkGhrqmOusZa3H43dr2FnkYHA/wwsuoARncJu00GzEGVrKAe1ZFCuS0ahzOTLinDsCYPgrhsMJgP+xyqqlLdWM1+93721e1je9V2dlTvwB/0oygKdY11VPmqqPZVE1C1+pxmJ8OTtGBiR/UOlh9YHhrBA5ASmRL68w6qQewmOxGGCDx+D26/+5h/J7s6ujI4cTBZjiwqvBU0NjeSE5NDsi2ZAw0HcDe5sZvs6BQdpZ5SPH4PTouTusY6lh9Yzvry9Yd9DS16Cz1cPYgwRLC7ZjcV3orQMavBSpYjiy6OLhQ3FLOmdM1hgd7BwCXCEEGCNYFKXyXuJq329Kh0Ls+9nPO6nteqv1Q/1vsbCTBa+Ru82tRE3RdfUDX7LXybNqGz23FOnoxL9zFGz9ZDw5ric7URF0cZYrTfvZ8JH07goZMe4qKeFx1xfHfNbibNncQfT/4jE7MmMu3baeTX5PPRpI9a9bUIIYRo/yTAED8mAYYQrU9VVYJq8GenB2yt3Mo+9z50io5kWzLZruxjXhMIBij1lFLoLqS4vhiH2UG2K5skW1Jo9ENjcyMrDqxgSdESTDoTSZFJmPQmAsEAe2r3sLVqK9uqtuENeA+7t9PspKax5mifNiQuIk4bLZA0jL5xfalvqqfUU0pJQwklDSWUekpxN7np6epJelQ6++r2saN6Bzurd1LmLTvifha9hURbIr1ie+EwOVhbtpZ6fz2X9LyEiV0nElSDGHVGHOZDK440B5uZVzCPLZVbiImIwaAYqGqsCoUVSbYkEm2J1DbWsrpsNfvr9lPVWAUq2E127GY7UaYoqnxVbK3UvhZHG0nzYxHBIGZVJQUj3c2xeIxm9ge97Pe7qQ82hs4z6Ux0i+6G1WDVwgeznRhLDNGWaGIsMfSP70+OK+ewP2NVVdnn3se2qm30ju1NSmQKAHVNdSwuXMya0jUEggFMehP94vrRN64vDpMDvU5Pg7+B4vpi1pStIa80j3Vl62jwN6BX9Bh0hlDI83NyXDkMT9aa3MZFxLG16lAQ1xhoJMuZRTdnN7o6u9LN2e2wv3MAld5KahpriDJFEWmMJMIQwfry9Xyy+xPcTW4cZgdZziwGxg+ke3T3/3kk0NFIgHEUrfkNPlBVRc2cOVT/458EyssxdemC67dX4Dj3XHQRFngqGQZfC2f+/LCazws+5+7Fd/Ovif8iJybniOP+Zj9D3h0SanK0uXIzKZEp/OPsf7TKaxFCCNFxSIAhfkwCDCE6l+ZgM0X1RVQ3VtPgb6CbsxtxEXFsq9rGosJFRBgiiIuII8oUhdVoxWa0YTfZj/ih9b/R1NxEIBhARaVZbUav6LEarGHvB1Ljq2FFyQoAuju742v2UVBbgF6nJ9YSS42vivwDq6mr3YO3voS9DSXsDjZgaw6Q5te29ECANL+ftIC2bzTaIDJOGzFviwNbLNhTwZ4EBguoKngqwVcLcT20xqXODGiFr0UgGKDaV43L4tKa0tbmU9pQSnJkMg6zg7qmOpqDzaHRMjWNNRgUA06L83/+3OF2rPc30gOjlRTfdx8NixZjO/lkkp56EtvIkSi6liSqYhcEfJCQe8zrZ66byeLCxQCUe8sx683HbIJi1Bu5uOfFrC9fT5WviiRbEmdkntHqr0kIIYQQQgjRvul1etLt6aSTftjzOTE5R/1laGsw6U3tculap8V5xM9FuTE/+hks8/QjLwwGwVcDniotjPBUgqcCGg5u5dBQBjX7oDBP+/inWGMhqS+YbKA3QVSS1pMjZSAk9v3FPTkMOgNx1rjQfo/oHvSI7hHaj42IPez8H++fiCTAaCXx06ah3HMP5qws7QlV1f4BABS1/BYk/ugBRpWvilc2vEKGPYPkyGRcFhcDEwb+5Koi9w27rzXLF0IIIYQQQojOSacDq0vb+AUraQQawV0CwYD2c5/Vpa2QUr4VilZD4Woo3aSd4/eC+4D2C20AvVlbRcWZoa06abGD2a49WhxgjdGWi3V1BUP7C4nCTQKMVmLp2fPwJz65DdbMDu0GFT1uexI01qKqKh/t+oh15esAbcRFQA3wzCnPHJaoCSGEEEIIIYRoZwxmiM448vmkfto2+JrDn1dVqCuCwlVawFFVANV7oWwrNNZCoxt+tNQvehMkD4DUIdrqKtGZkD5MCz06MQkw2sqe77S/cH0vAeDBqhV88uHhw5Uy7ZmhdaXP73a+hBdCCCGEEEIIcaJRFHCkaluv8488rqrQVA++Om16SsUuKNkA+5bDylcg1LxTgZiu2tQUsx1cWdo9DZaWzXzoUW3Wpr80N0GEC+zJ2vQVW8xxfemtTQKMttDohqp8GPMgnHQDAGv+/RF9YvtwdtbZgLZm9OjU0WFvdCOEEEIIIYQQIowUBcxR2uZIaflF+IXasYOtCcq3ab8kL92sTWHxVsG2z7ReHf8NUxSYrGCMAKNNWxVTDWoBSkOFNhUmeQDEdtOCD50emhq0kR/xOdrUlqhE7V4N5VCySRtZktQXcs5p3a/LUUiA0RZKt2iPib0B8Pg9FNUXManbJC7LuSyMhQkhhBBCCCGE6DAURRs1YRsJmSOPPN7s1wKNQKPWZ+Pgpui1VVP0Bi0AqdkLBzZAXTH4PVpvDr9H6+Oh6LVQwxanNTAtWgP53xzq23FkUfDD5WoVHQy/RQKMDuWrR2DXAu1jX632mNgHgD11ewDo6uwahsKEEEIIIYQQQpyQ9EZtM0ce+xyLQ2scmnXqf3fvJg+ggiFCG+lRulkLQmqLQGfQlpSN7aGN2Pipz9+KJMBoLbY4rbFKi4bss9nZWAFlFXx/4HsAujokwBBCCCGEEEII0QGYrIc+jozXtjCTAKO1jLhF21r836K7+fzzK0L7EYYI0uxp4ahMCCGEEEIIIYTo8CTAaCN76vbQN7YvN/W/CYAkWxJGnTHMVQkhhBBCCCGEEB2TBBhtpNRTytj0sYxMOUqjFSGEEEIIIYQQQvxXdOEu4ETU2NxIla+KBGtCuEsRQgghhBBCCCFOCBJgtIGyhjIAEm2JYa5ECCGEEEIIIYQ4MUiA0QZKPCUAMgJDCCGEEEIIIYRoJdIDo5V4A14CwQAA++r2AZBgkwBDCCGEEEIIIYRoDRJgtJJHlj3C5wWfh/YVFBKtMoVECCGEEEIIIYRoDRJgtJKJWRPpHdM7tJ8alYrVaA1jRUIIIYQQQgghxIlDAoxWMjp1NKNTR4e7DCGEEEIIIYQQ4oTU7pt4KoqSrSjKckVRGhVFufMnzuuiKMoKRVF2KooyR1EU0/GsUwghhBBCCCGEEG2n3QcYQBVwK/Dsz5z3NPC8qqrdgWrg2rYuTAghhBBCCCGEEMdHuw8wVFUtU1V1FeA/1jmKoijAWOCDlqdmA5OOQ3lCCCGEEEIIIYQ4Dtp9gPELxQA1qqoGWvYLgZSjnagoyvWKouQpipJXXl5+3AoUQgghhBBCCCHEr3eiBBjKUZ5Tj3aiqqovq6o6WFXVwXFxcW1clhBCCCGEEEIIIVpDuwwwFEW5WVGUdS1b8i+4pAJwKopycFWVVKC47SoUQgghhBBCCCHE8dQuAwxVVWeoqtq/ZfvZIEJVVRX4Bpjc8tSVwNy2rFEIIYQQQgghhBDHT7sMMH5IUZRERVEKgWnAg4qiFCqKYm85Nu8HIzTuAaYpirILrSfGa+GpWAghhBCi9SiKcqaiKNsVRdmlKMq94a5HCCGECBfDz58SXqqqlqBNCTnasQk/+DgfGHq86hJCCCGEaGuKouiBGcB4tCblqxRF+VhV1S3hrUwIIYQ4/tr9CAwhhBBCiE5sKLBLVdX8/2/v7oPlmu84jr8/TSQSQURESVKSNkxV0whSGs8MgolOazwMg9FORkerYlSjQeW/ooPBVBnPpUk9puqhjZIwVUnkWeIhgiBEw9Qz4yH99o/zu5N17U323ru753dzP6+ZM3v2t2d3P/u9u+ec+9vfORsRnwHTgKNLzmRmZlYKd2CYmZmZ5Wsw8FrF9TZ/Kt7MzGxjl/0hJI00f/78tyW9UseHHEjxiyjWfq5dx7l2HefadZxr1zn1rN8OdXocy1NNPxUvaQIwIV39UNLznXze3D/jOefLORs4X2fknA3yzpdzNsg7X87ZoHH5qu7fdOsOjIjYpp6PJ2leROxRz8fsLly7jnPtOs616zjXrnNcP2uHVcDQiutVfyo+Iq4DrqvXk+b+Hs05X87ZwPk6I+dskHe+nLNB3vlyzgbNz+dDSMzMzMzy9RQwQtIwSb2A44H7Ss5kZmZWim49AsPMzMwsZxHxhaSfA/8AegA3RsSykmOZmZmVwh0Y9VW3oZvdkGvXca5dx7l2HefadY7rZzWLiAeBB5v8tLm/R3POl3M2cL7OyDkb5J0v52yQd76cs0GT8yniK+eBMjMzMzMzMzPLis+BYWZmZmZmZmbZcwdGHUi6UdIaSUvLzpKjavWRNEDSw5JeSJdbVdx2nqQVkp6XdFg5qfMgaaikmZKelbRM0i9Tu+u3AZI2lTRX0uJUuymp3bWrkaQekhZKuj9dd+1qJGmlpKclLZI0L7W5ftYlSDo8vRdXSJpUwvNnv+3Lef0oqb+kuyQ9l2q4d2b5Jqa/61JJU9P2upR89dpHlbR7WuevkHSlpGo/f1yvfJemv+0SSfdK6p9TvorbzpEUkgaWka+tbJJ+kZ5/maRLysjWVj5JoyTNVtp3kDSmjHyq4zq4IfWLCE+dnID9gNHA0rKz5DhVqw9wCTApzU8CLk7zuwCLgd7AMOBFoEfZr6HE2m0HjE7zmwPLU41cvw3XTkC/NL8JMAfYy7VrVw3PBv4M3J+uu3a1124lMLBVm+vnKfuJ4kShLwLDgV7pvblLkzNkv+3Lef0I3AL8NM33Avrnkg8YDLwM9EnX7wBOLSsfddpHBeYCe1PsezwEjGtgvkOBnmn+4tzypfahFCcefoW0LWx2vjZqdyDwT6B3uj4op9oBM1oeHzgCmFVS7eq2Dm5EPo/AqIOIeBz4b9k5ctVGfY6m2MCSLn9Y0T4tIj6NiJeBFcAYuqmIWB0RC9L8B8CzFBt/128DovBhurpJmgLXriaShgBHAtdXNLt2neP6WVcwBlgRES9FxGfARDVQSAAACFdJREFUNIr3aNPkvu3Lef0oaQuKf4xuAIiIzyLi3VzyJT2BPpJ6An2BN8rKV499VEnbAVtExJNR/Md2a8V96p4vImZExBfp6mxgSE75ksuBcyn2u1o0NV8b2X4G/C4iPk3LrCkj23ryBbBFmt+S4rPR9Hz1Wgc3Kp87MKws20bEaig+JMCg1D4YeK1iuVWprduTtCOwG8VIAtevBiqG+C4C1gAPR4RrV7srKHY+/lfR5trVLoAZkuZLmpDaXD/rCrJ6P2a67ct5/TgceAu4ScUhLtdL2iyXfBHxOvB74FVgNfBeRMzIJV/S3iyD03wzM7Y4jeJbbcgkn6TxwOsRsbjVTTnk2wnYV9IcSY9J2jOjbABnAZdKeo3ic3Je2fk6uQ5uSD53YFhuqh0X1e1/KkdSP+Bu4KyIeH99i1Zp67b1i4i1ETGK4tuJMZJ2Xc/irl0i6ShgTUTMr/UuVdq6Ze0qjI2I0cA44AxJ+61nWdfPcpLN+zHHbV8XWD/2pBiWfk1E7AZ8RDHUuy1NzZeOmT+aYpj59sBmkk5a312qtJW1fmwrSykZJU0GvgBub2lqI0czPx99gcnAhdVubiNHM+vXE9iK4pDiXwF3pHMy5JANihEiEyNiKDCRNJJqPTkamq8O6+CG5HMHhpXlP2lYEemyZQjXKorj5loMYd3wqW5J0iYUK4/bI+Ke1Oz6tUMaPjsLOBzXrhZjgfGSVlIMHz9I0m24djWLiDfS5RrgXoohz66fdQVZvB8z3vblvn5cBaxKIw4B7qLo0Mgl3yHAyxHxVkR8DtwD/CCjfHQgyyrWHcbRlIySTgGOAk5MQ/NzyfdNis6pxekzMgRYIOnrmeRbBdyTDjOeSzGKamAm2QBOofhMANzJusOlmp6vTuvghuRzB4aV5T6KDynp8q8V7cdL6i1pGDCC4uQv3VLqFb4BeDYiLqu4yfXbAEnbKJ2ZW1Ifip2m53DtNigizouIIRGxI3A88GhEnIRrVxNJm0navGWe4oRrS3H9rGt4ChghaZikXhTrgPuaGSDnbV/u68eIeBN4TdLOqelg4Jlc8lEcOrKXpL7p73wwxfH1ueRrec6as6Sh9B9I2iu9ppMr7lN3kg4Hfg2Mj4iPW+UuNV9EPB0RgyJix/QZWUVxMsg3c8gHTAcOApC0E8VJbt/OJBsU/9zvn+YPAl5I803NV691cMPqF3U4i2p3n4CpFMfxfU7xQf1J2ZlymqrVB9gaeITig/kIMKBi+ckUZ699njqd6berTsA+FEOtlgCL0nSE61dT7UYCC1PtlgIXpnbXrn11PIB1Z9l37Wqr2XCKs3EvBpYBk10/T11pStuZ5ek9ObmE5+8S275c14/AKGBeqt90iiHzOeWbQvGFwlLgTxS/XFBKPuq0jwrskV7Pi8DVgBqYbwXF+QZaPht/zClfq9tXUvGLXM3M10btegG3pedaAByUU+0o1n3zKfYf5gC7l1S7uq2DG5FP6YHNzMzMzMzMzLLlQ0jMzMzMzMzMLHvuwDAzMzMzMzOz7LkDw8zMzMzMzMyy5w4MMzMzMzMzM8ueOzDMzMzMzMzMLHvuwDCzdpE0WdIySUskLZL0fUlnSepbw31rWs7MzMysHiStTfsrLdOkDSx/uqST6/C8KyUN7OzjmNmX+WdUzaxmkvYGLgMOiIhP04a5F/BvYI+IeHsD919Zy3JmZmZm9SDpw4joV8LzrsT7PGZ15xEYZtYe2wFvR8SnAGmjfAywPTBT0kwASddImpdGakxJbWdWWe5QSU9KWiDpTklN38EwMzOz7ieNkLhY0tw0fSu1XyTpnDR/pqRn0qjTaaltgKTpqW22pJGpfWtJMyQtlHQtoIrnOik9xyJJ10rqkaabJS2V9LSkiSWUwazLcQeGmbXHDGCopOWS/iBp/4i4EngDODAiDkzLTY6IPYCRwP6SRrZeLo3eOB84JCJGA/OAs5v/kszMzGwj1qfVISTHVdz2fkSMAa4Grqhy30nAbhExEjg9tU0BFqa23wC3pvbfAv+KiN2A+4BvAEj6NnAcMDYiRgFrgROBUcDgiNg1Ir4L3FTH12y20epZdgAz6zoi4kNJuwP7AgcCf2njWNJjJU2gWMdsB+wCLGm1zF6p/QlJUByK8mSjspuZmVm39EnqOKhmasXl5VVuXwLcLmk6MD217QP8GCAiHk0jL7YE9gN+lNofkPROWv5gYHfgqbS/0wdYA/wNGC7pKuABii+JzGwD3IFhZu0SEWuBWcAsSU8Dp1TeLmkYcA6wZ0S8I+lmYNMqDyXg4Yg4obGJzczMzKqKNuZbHEnRMTEeuEDSd6g4NKTKfas9hoBbIuK8r9wgfQ84DDgDOBY4rfboZt2TDyExs5pJ2lnSiIqmUcArwAfA5qltC+Aj4D1J2wLjKpavXG42MLbimNO+knZqZH4zMzOzCsdVXH5pFKikrwFDI2ImcC7QH+gHPE5xCAiSDqA4N9j7rdrHAVulh3oEOEbSoHTbAEk7pENpvxYRdwMXAKMb9SLNNiYegWFm7dEPuEpSf+ALYAUwATgBeEjS6nR+i4XAMuAl4ImK+1/XarlTgamSeqfbzweWN+m1mJmZ2cavj6RFFdf/HhEth7/2ljSH4kvd1iNCewC3pcNDBFweEe9Kugi4SdIS4GPWjUSdQrFPswB4DHgVICKekXQ+MCN1inxOMeLik/Q4LV8of2WEhpl9lX9G1czMzMzMuhX/zKlZ1+RDSMzMzMzMzMwsex6BYWZmZmZmZmbZ8wgMMzMzMzMzM8ueOzDMzMzMzMzMLHvuwDAzMzMzMzOz7LkDw8zMzMzMzMyy5w4MMzMzMzMzM8ueOzDMzMzMzMzMLHv/Bwns4hlMSaK0AAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 1080x360 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x360 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# ---------------\n",
    "# Discussion Cell\n",
    "# ---------------\n",
    "\n",
    "# Make sure to verify your experiment result with the test cell above.\n",
    "# Otherwise the sweep results will not be displayed.\n",
    "\n",
    "# Experiment parameters\n",
    "experiment_parameters = {\n",
    "    \"num_runs\" : 50,\n",
    "    \"num_episodes\" : 2000,\n",
    "    \"episode_eval_frequency\" : 10 # evaluate every 10 episodes\n",
    "}\n",
    "\n",
    "# Environment parameters\n",
    "environment_parameters = {\n",
    "    \"num_states\" : 500,\n",
    "    \"start_state\" : 250,\n",
    "    \"left_terminal_state\" : 0,\n",
    "    \"right_terminal_state\" : 501,\n",
    "    \"discount_factor\" : 1.0\n",
    "}\n",
    "\n",
    "# Agent parameters\n",
    "# Each element is an array because we will be sweeping over multiple values\n",
    "agent_parameters = {\n",
    "    \"num_groups\": [10, 100, 500],\n",
    "    \"step_size\": [0.01, 0.05, 0.1]\n",
    "}\n",
    "\n",
    "if all_correct:\n",
    "    plot_script.plot_result(agent_parameters, 'correct_npy')\n",
    "else:\n",
    "    raise ValueError(\"Make sure your experiment result is correct! Otherwise the sweep results will not be displayed.\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "markdown",
     "checksum": "f781480c3fab83e00a9edad13f1a1d2d",
     "grade": false,
     "grade_id": "cell-97905e6ab8fc9184",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "source": [
    "## Wrapping up\n",
    "\n",
    "Let’s think about the results of our parameter study.\n",
    "\n",
    "### State Aggregation\n",
    "\n",
    "- Which state aggregation resolution do you think is the best after running 2000 episodes? Which state aggregation resolution do you think would be the best if we could train for only 200 episodes? What if we could train for a million episodes?\n",
    "\n",
    "- Should we use tabular representation (state aggregation of resolution 500) whenever possible? Why might we want to use function approximation?\n",
    "\n",
    "\n",
    "From the plots, using 100 state aggregation with step-size 0.05 reaches the best performance: the lowest RMSVE after 2000 episodes. If the agent can only be trained for 200 episodes, then 10 state aggregation with step-size 0.05 reaches the lowest error. Increasing the resolution of state aggregation makes the function approximation closer to a  tabular representation, which would be able to learn exactly correct state values for all states. But learning will be slower. \n",
    "\n",
    "\n",
    "### Step-Size\n",
    "\n",
    "- How did different step-sizes affect learning?\n",
    "\n",
    "The best step-size is different for different state aggregation resolutions. A larger step-size allows the agent to learn faster, but might not perform as well asymptotically. A smaller step-size causes it to learn more slowly, but may perform well asymptotically."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "markdown",
     "checksum": "88ec1cc6e29705e4b6524a8f9d9a8b00",
     "grade": false,
     "grade_id": "cell-7a94c1ffa89569ce",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "source": [
    "### **Congratulations!** You have successfully implemented Course 3 Programming Assignment 1.\n",
    "\n",
    "You have implemented **semi-gradient TD(0) with State Aggregation** in a 500-state Random Walk. We used an environment with a large but discrete state space, where it was possible to compute the true state values. This allowed us to compare the values learned by your agent to the true state values. The same state aggregation function approximation can also be applied to continuous state space environments, where comparison to the true values is not usually possible.\n",
    "\n",
    "\n",
    "You also successfully applied supervised learning approaches to approximate value functions with semi-gradient TD(0). \n",
    "\n",
    "Finally, we plotted the learned state values and compared with true state values. We also compared learning curves of different state aggregation resolutions and learning rates. \n",
    "\n",
    "From the results, you can  see why it is often desirable to use function approximation, even when tabular learning is possible. Asymptotically, an agent with tabular representation would be able to learn the true state value function, but it would learn much more slowly compared to an agent with function approximation. On the other hand, we also want to ensure we do not reduce discrimination too far (a coarse state aggregation resolution), because it will hurt the asymptotic performance.\n"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}