Example.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "%load_ext autoreload\n",
    "%autoreload 2\n",
    "\n",
    "# Imports\n",
    "## Common libs\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import pickle\n",
    "from itertools import cycle, chain\n",
    "from copy import deepcopy\n",
    "from datetime import datetime\n",
    "from collections import OrderedDict\n",
    "\n",
    "## Plotting\n",
    "import seaborn as sns\n",
    "from matplotlib import pyplot as plt\n",
    "from matplotlib.ticker import ScalarFormatter\n",
    "from matplotlib.pyplot import rc\n",
    "\n",
    "## Graphs\n",
    "from networkx.drawing import nx_agraph\n",
    "import networkx as nx\n",
    "from graphviz import Source\n",
    "import pygraphviz\n",
    "\n",
    "## GPs\n",
    "from GPy.kern import RBF\n",
    "from GPy.models.gp_regression import GPRegression\n",
    "\n",
    "## DCBO Utils\n",
    "import sys\n",
    "sys.path.append(\"../DCBO\")\n",
    "from src.utils.utilities import powerset\n",
    "from src.utils.sem_utils.sem_estimate import build_sem_hat\n",
    "from src.utils.sequential_intervention_functions import get_interventional_grids\n",
    "from src.experimental.experiments import optimal_sequence_of_interventions, run_methods_replicates\n",
    "from src.experimental.analyse_results import get_relevant_results, elaborate\n",
    "from src.utils.plotting import plot_expected_opt_curve_paper\n",
    "\n",
    "## Optimisation Algos\n",
    "from src.methods.cbo import CBO\n",
    "from src.methods.dcbo import DCBO\n",
    "from src.methods.bo import BO\n",
    "\n",
    "## YT\n",
    "import random\n",
    "from gym import spaces\n",
    "from numpy import repeat\n",
    "from stable_baselines3.common.env_checker import check_env\n",
    "\n",
    "sys.path.append(\"../YAWNING-TITAN_fork\")\n",
    "from yawning_titan.integrations.dcbo.dcbo_agent import DCBOAgent\n",
    "from yawning_titan.envs.generic.core.blue_interface import BlueInterface\n",
    "from yawning_titan.envs.generic.core.network_interface import NetworkInterface\n",
    "from yawning_titan.envs.generic.core.red_interface import RedInterface\n",
    "from yawning_titan.envs.generic.generic_env import GenericNetworkEnv\n",
    "from yawning_titan.envs.generic.helpers import network_creator"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Problem Setup\n",
    "* We want to use YAWNING TITAN to simulate data for a range of probabilities for $P$ and $I$\n",
    "* We want to generate data for 50 timesteps in each environment\n",
    "* We then want to use BO, CBO and DCBO to take that data and determine the optimal value for $P$ and $I$ in the final 3 timesteps"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### First, we need to define our DAG\n",
    "* This will be our causal model of how compromise spreads in our example network (Fig. 2 in the paper)\n",
    "* It will also define what data we need to collect from YAWNING TITAN"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# How many timesteps we want to visaulise:\n",
    "T = 3"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "def picshat_graph(start_time: int, stop_time: int, verbose=False):\n",
    "    \"\"\"\n",
    "    Helper function to construct our DAG as a NetworkX graph.\n",
    "    \"\"\"\n",
    "    assert start_time <= stop_time\n",
    "\n",
    "    spatial_edges, ranking = [], []\n",
    "    nodes = [\"P\", \"I\", \"H\", \"A\", \"C\", \"T\", \"S\"]\n",
    "    node_count = len(nodes)\n",
    "    connections = node_count * \"{}_{} -> {}_{}; \"\n",
    "    edge_pairs = [\"P\", \"A\", \"P\", \"H\", \"I\", \"A\", \"I\", \"S\", \"C\", \"H\", \"C\", \"T\", \"A\", \"T\"]\n",
    "    pair_count = len(edge_pairs)\n",
    "    for t in range(start_time, stop_time + 1):\n",
    "        space_idx = pair_count * [t]\n",
    "        iters = [iter(edge_pairs), iter(space_idx)]\n",
    "        inserts = list(chain(map(next, cycle(iters)), *iters))\n",
    "        spatial_edges.append(connections.format(*inserts))\n",
    "        ranking.append(\n",
    "            \"{{ rank=same; {} }} \".format(\n",
    "                \" \".join([item + \"_{}\".format(t) for item in nodes])\n",
    "            )\n",
    "        )\n",
    "    ranking = \"\".join(ranking)\n",
    "    spatial_edges = \"\".join(spatial_edges)\n",
    "\n",
    "    temporal_edges = []\n",
    "    NUMBER_TRANS_EDGES = 2\n",
    "    connections = NUMBER_TRANS_EDGES * \"{}_{} -> {}_{}; \"\n",
    "    for t in range(stop_time):\n",
    "        edge_pairs = [\"C\", \"C\", \"S\", \"S\"]\n",
    "        temporal_idx = node_count * [t, t + 1]\n",
    "        iters = [iter(edge_pairs), iter(temporal_idx)]\n",
    "        inserts = list(chain(map(next, cycle(iters)), *iters))\n",
    "        temporal_edges.append(connections.format(*inserts))\n",
    "    temporal_edges.append(\"\")\n",
    "    temporal_edges = \"\".join(temporal_edges)\n",
    "    graph = \"digraph {{ rankdir=LR; {} {} {} }}\".format(\n",
    "        spatial_edges, temporal_edges, ranking\n",
    "    )\n",
    "    if verbose:\n",
    "        return Source(graph)\n",
    "    return graph\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "dag_view = picshat_graph(0, T - 1, verbose=True)\n",
    "dag = nx_agraph.from_agraph(pygraphviz.AGraph(dag_view.source))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Let's look at our DAG to make sure it follows Figure 2 of the paper:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/svg+xml": "<?xml version=\"1.0\" encoding=\"UTF-8\" standalone=\"no\"?>\n<!DOCTYPE svg PUBLIC \"-//W3C//DTD SVG 1.1//EN\"\n \"http://www.w3.org/Graphics/SVG/1.1/DTD/svg11.dtd\">\n<!-- Generated by graphviz version 2.43.0 (0)\n -->\n<!-- Title: %3 Pages: 1 -->\n<svg width=\"284pt\" height=\"462pt\"\n viewBox=\"0.00 0.00 284.13 462.00\" xmlns=\"http://www.w3.org/2000/svg\" xmlns:xlink=\"http://www.w3.org/1999/xlink\">\n<g id=\"graph0\" class=\"graph\" transform=\"scale(1 1) rotate(0) translate(4 458)\">\n<title>%3</title>\n<polygon fill=\"white\" stroke=\"transparent\" points=\"-4,4 -4,-458 280.13,-458 280.13,4 -4,4\"/>\n<!-- P_0 -->\n<g id=\"node1\" class=\"node\">\n<title>P_0</title>\n<ellipse fill=\"none\" stroke=\"black\" cx=\"61.14\" cy=\"-436\" rx=\"27.1\" ry=\"18\"/>\n<text text-anchor=\"middle\" x=\"61.14\" y=\"-432.3\" font-family=\"Times,serif\" font-size=\"14.00\">P_0</text>\n</g>\n<!-- A_0 -->\n<g id=\"node2\" class=\"node\">\n<title>A_0</title>\n<ellipse fill=\"none\" stroke=\"black\" cx=\"61.14\" cy=\"-166\" rx=\"27.1\" ry=\"18\"/>\n<text text-anchor=\"middle\" x=\"61.14\" y=\"-162.3\" font-family=\"Times,serif\" font-size=\"14.00\">A_0</text>\n</g>\n<!-- P_0&#45;&gt;A_0 -->\n<g id=\"edge1\" class=\"edge\">\n<title>P_0&#45;&gt;A_0</title>\n<path fill=\"none\" stroke=\"black\" d=\"M39.27,-425.03C29.83,-419.03 19.77,-410.64 14.55,-400 -4.85,-360.51 -4.85,-241.49 14.55,-202 18.22,-194.52 24.29,-188.15 30.84,-182.94\"/>\n<polygon fill=\"black\" stroke=\"black\" points=\"33.13,-185.61 39.27,-176.97 29.09,-179.9 33.13,-185.61\"/>\n</g>\n<!-- H_0 -->\n<g id=\"node3\" class=\"node\">\n<title>H_0</title>\n<ellipse fill=\"none\" stroke=\"black\" cx=\"61.14\" cy=\"-328\" rx=\"28.7\" ry=\"18\"/>\n<text text-anchor=\"middle\" x=\"61.14\" y=\"-324.3\" font-family=\"Times,serif\" font-size=\"14.00\">H_0</text>\n</g>\n<!-- P_0&#45;&gt;H_0 -->\n<g id=\"edge2\" class=\"edge\">\n<title>P_0&#45;&gt;H_0</title>\n<path fill=\"none\" stroke=\"black\" d=\"M39.27,-425.03C29.83,-419.03 19.77,-410.64 14.55,-400 7.49,-385.64 7.49,-378.36 14.55,-364 18.16,-356.64 24.08,-350.37 30.5,-345.21\"/>\n<polygon fill=\"black\" stroke=\"black\" points=\"32.69,-347.95 38.78,-339.29 28.61,-342.26 32.69,-347.95\"/>\n</g>\n<!-- T_0 -->\n<g id=\"node7\" class=\"node\">\n<title>T_0</title>\n<ellipse fill=\"none\" stroke=\"black\" cx=\"61.14\" cy=\"-112\" rx=\"27\" ry=\"18\"/>\n<text text-anchor=\"middle\" x=\"61.14\" y=\"-108.3\" font-family=\"Times,serif\" font-size=\"14.00\">T_0</text>\n</g>\n<!-- A_0&#45;&gt;T_0 -->\n<g id=\"edge7\" class=\"edge\">\n<title>A_0&#45;&gt;T_0</title>\n<path fill=\"none\" stroke=\"black\" d=\"M61.14,-147.79C61.14,-145.31 61.14,-142.83 61.14,-140.34\"/>\n<polygon fill=\"black\" stroke=\"black\" points=\"64.64,-140.14 61.14,-130.14 57.64,-140.14 64.64,-140.14\"/>\n</g>\n<!-- I_0 -->\n<g id=\"node4\" class=\"node\">\n<title>I_0</title>\n<ellipse fill=\"none\" stroke=\"black\" cx=\"61.14\" cy=\"-274\" rx=\"27\" ry=\"18\"/>\n<text text-anchor=\"middle\" x=\"61.14\" y=\"-270.3\" font-family=\"Times,serif\" font-size=\"14.00\">I_0</text>\n</g>\n<!-- I_0&#45;&gt;A_0 -->\n<g id=\"edge3\" class=\"edge\">\n<title>I_0&#45;&gt;A_0</title>\n<path fill=\"none\" stroke=\"black\" d=\"M39.27,-263.03C29.83,-257.03 19.77,-248.64 14.55,-238 7.49,-223.64 7.49,-216.36 14.55,-202 18.22,-194.52 24.29,-188.15 30.84,-182.94\"/>\n<polygon fill=\"black\" stroke=\"black\" points=\"33.13,-185.61 39.27,-176.97 29.09,-179.9 33.13,-185.61\"/>\n</g>\n<!-- S_0 -->\n<g id=\"node5\" class=\"node\">\n<title>S_0</title>\n<ellipse fill=\"none\" stroke=\"black\" cx=\"61.14\" cy=\"-220\" rx=\"27.1\" ry=\"18\"/>\n<text text-anchor=\"middle\" x=\"61.14\" y=\"-216.3\" font-family=\"Times,serif\" font-size=\"14.00\">S_0</text>\n</g>\n<!-- I_0&#45;&gt;S_0 -->\n<g id=\"edge4\" class=\"edge\">\n<title>I_0&#45;&gt;S_0</title>\n<path fill=\"none\" stroke=\"black\" d=\"M61.14,-255.79C61.14,-253.31 61.14,-250.83 61.14,-248.34\"/>\n<polygon fill=\"black\" stroke=\"black\" points=\"64.64,-248.14 61.14,-238.14 57.64,-248.14 64.64,-248.14\"/>\n</g>\n<!-- S_1 -->\n<g id=\"node12\" class=\"node\">\n<title>S_1</title>\n<ellipse fill=\"none\" stroke=\"black\" cx=\"154.34\" cy=\"-234\" rx=\"27.1\" ry=\"18\"/>\n<text text-anchor=\"middle\" x=\"154.34\" y=\"-230.3\" font-family=\"Times,serif\" font-size=\"14.00\">S_1</text>\n</g>\n<!-- S_0&#45;&gt;S_1 -->\n<g id=\"edge23\" class=\"edge\">\n<title>S_0&#45;&gt;S_1</title>\n<path fill=\"none\" stroke=\"black\" d=\"M88.03,-223.97C97.26,-225.39 107.81,-227.01 117.71,-228.53\"/>\n<polygon fill=\"black\" stroke=\"black\" points=\"117.29,-232.01 127.7,-230.06 118.35,-225.09 117.29,-232.01\"/>\n</g>\n<!-- C_0 -->\n<g id=\"node6\" class=\"node\">\n<title>C_0</title>\n<ellipse fill=\"none\" stroke=\"black\" cx=\"61.14\" cy=\"-382\" rx=\"27.9\" ry=\"18\"/>\n<text text-anchor=\"middle\" x=\"61.14\" y=\"-378.3\" font-family=\"Times,serif\" font-size=\"14.00\">C_0</text>\n</g>\n<!-- C_0&#45;&gt;H_0 -->\n<g id=\"edge5\" class=\"edge\">\n<title>C_0&#45;&gt;H_0</title>\n<path fill=\"none\" stroke=\"black\" d=\"M61.14,-363.79C61.14,-361.31 61.14,-358.83 61.14,-356.34\"/>\n<polygon fill=\"black\" stroke=\"black\" points=\"64.64,-356.14 61.14,-346.14 57.64,-356.14 64.64,-356.14\"/>\n</g>\n<!-- C_0&#45;&gt;T_0 -->\n<g id=\"edge6\" class=\"edge\">\n<title>C_0&#45;&gt;T_0</title>\n<path fill=\"none\" stroke=\"black\" d=\"M38.78,-370.71C29.49,-364.74 19.68,-356.46 14.55,-346 -4.85,-306.51 -4.85,-187.49 14.55,-148 18.22,-140.52 24.29,-134.15 30.84,-128.94\"/>\n<polygon fill=\"black\" stroke=\"black\" points=\"33.13,-131.61 39.27,-122.97 29.09,-125.9 33.13,-131.61\"/>\n</g>\n<!-- C_1 -->\n<g id=\"node13\" class=\"node\">\n<title>C_1</title>\n<ellipse fill=\"none\" stroke=\"black\" cx=\"154.34\" cy=\"-355\" rx=\"27.9\" ry=\"18\"/>\n<text text-anchor=\"middle\" x=\"154.34\" y=\"-351.3\" font-family=\"Times,serif\" font-size=\"14.00\">C_1</text>\n</g>\n<!-- C_0&#45;&gt;C_1 -->\n<g id=\"edge22\" class=\"edge\">\n<title>C_0&#45;&gt;C_1</title>\n<path fill=\"none\" stroke=\"black\" d=\"M87.06,-374.62C96.89,-371.71 108.35,-368.32 118.95,-365.18\"/>\n<polygon fill=\"black\" stroke=\"black\" points=\"120.17,-368.47 128.77,-362.28 118.18,-361.76 120.17,-368.47\"/>\n</g>\n<!-- P_1 -->\n<g id=\"node8\" class=\"node\">\n<title>P_1</title>\n<ellipse fill=\"none\" stroke=\"black\" cx=\"154.34\" cy=\"-180\" rx=\"27.1\" ry=\"18\"/>\n<text text-anchor=\"middle\" x=\"154.34\" y=\"-176.3\" font-family=\"Times,serif\" font-size=\"14.00\">P_1</text>\n</g>\n<!-- A_1 -->\n<g id=\"node9\" class=\"node\">\n<title>A_1</title>\n<ellipse fill=\"none\" stroke=\"black\" cx=\"154.34\" cy=\"-72\" rx=\"27.1\" ry=\"18\"/>\n<text text-anchor=\"middle\" x=\"154.34\" y=\"-68.3\" font-family=\"Times,serif\" font-size=\"14.00\">A_1</text>\n</g>\n<!-- P_1&#45;&gt;A_1 -->\n<g id=\"edge8\" class=\"edge\">\n<title>P_1&#45;&gt;A_1</title>\n<path fill=\"none\" stroke=\"black\" d=\"M132.46,-169.03C123.03,-163.03 112.97,-154.64 107.74,-144 100.69,-129.64 100.69,-122.36 107.74,-108 111.42,-100.52 117.48,-94.15 124.04,-88.94\"/>\n<polygon fill=\"black\" stroke=\"black\" points=\"126.33,-91.61 132.46,-82.97 122.28,-85.9 126.33,-91.61\"/>\n</g>\n<!-- H_1 -->\n<g id=\"node10\" class=\"node\">\n<title>H_1</title>\n<ellipse fill=\"none\" stroke=\"black\" cx=\"154.34\" cy=\"-126\" rx=\"28.7\" ry=\"18\"/>\n<text text-anchor=\"middle\" x=\"154.34\" y=\"-122.3\" font-family=\"Times,serif\" font-size=\"14.00\">H_1</text>\n</g>\n<!-- P_1&#45;&gt;H_1 -->\n<g id=\"edge9\" class=\"edge\">\n<title>P_1&#45;&gt;H_1</title>\n<path fill=\"none\" stroke=\"black\" d=\"M154.34,-161.79C154.34,-159.31 154.34,-156.83 154.34,-154.34\"/>\n<polygon fill=\"black\" stroke=\"black\" points=\"157.84,-154.14 154.34,-144.14 150.84,-154.14 157.84,-154.14\"/>\n</g>\n<!-- T_1 -->\n<g id=\"node14\" class=\"node\">\n<title>T_1</title>\n<ellipse fill=\"none\" stroke=\"black\" cx=\"154.34\" cy=\"-18\" rx=\"27\" ry=\"18\"/>\n<text text-anchor=\"middle\" x=\"154.34\" y=\"-14.3\" font-family=\"Times,serif\" font-size=\"14.00\">T_1</text>\n</g>\n<!-- A_1&#45;&gt;T_1 -->\n<g id=\"edge14\" class=\"edge\">\n<title>A_1&#45;&gt;T_1</title>\n<path fill=\"none\" stroke=\"black\" d=\"M154.34,-53.79C154.34,-51.31 154.34,-48.83 154.34,-46.34\"/>\n<polygon fill=\"black\" stroke=\"black\" points=\"157.84,-46.14 154.34,-36.14 150.84,-46.14 157.84,-46.14\"/>\n</g>\n<!-- I_1 -->\n<g id=\"node11\" class=\"node\">\n<title>I_1</title>\n<ellipse fill=\"none\" stroke=\"black\" cx=\"154.34\" cy=\"-288\" rx=\"27\" ry=\"18\"/>\n<text text-anchor=\"middle\" x=\"154.34\" y=\"-284.3\" font-family=\"Times,serif\" font-size=\"14.00\">I_1</text>\n</g>\n<!-- I_1&#45;&gt;A_1 -->\n<g id=\"edge10\" class=\"edge\">\n<title>I_1&#45;&gt;A_1</title>\n<path fill=\"none\" stroke=\"black\" d=\"M132.46,-277.03C123.03,-271.03 112.97,-262.64 107.74,-252 93.64,-223.28 93.64,-136.72 107.74,-108 111.42,-100.52 117.48,-94.15 124.04,-88.94\"/>\n<polygon fill=\"black\" stroke=\"black\" points=\"126.33,-91.61 132.46,-82.97 122.28,-85.9 126.33,-91.61\"/>\n</g>\n<!-- I_1&#45;&gt;S_1 -->\n<g id=\"edge11\" class=\"edge\">\n<title>I_1&#45;&gt;S_1</title>\n<path fill=\"none\" stroke=\"black\" d=\"M154.34,-269.79C154.34,-267.31 154.34,-264.83 154.34,-262.34\"/>\n<polygon fill=\"black\" stroke=\"black\" points=\"157.84,-262.14 154.34,-252.14 150.84,-262.14 157.84,-262.14\"/>\n</g>\n<!-- S_2 -->\n<g id=\"node19\" class=\"node\">\n<title>S_2</title>\n<ellipse fill=\"none\" stroke=\"black\" cx=\"247.53\" cy=\"-193\" rx=\"27.1\" ry=\"18\"/>\n<text text-anchor=\"middle\" x=\"247.53\" y=\"-189.3\" font-family=\"Times,serif\" font-size=\"14.00\">S_2</text>\n</g>\n<!-- S_1&#45;&gt;S_2 -->\n<g id=\"edge25\" class=\"edge\">\n<title>S_1&#45;&gt;S_2</title>\n<path fill=\"none\" stroke=\"black\" d=\"M177.41,-224.08C188.84,-218.94 202.98,-212.58 215.49,-206.96\"/>\n<polygon fill=\"black\" stroke=\"black\" points=\"216.98,-210.13 224.66,-202.83 214.11,-203.74 216.98,-210.13\"/>\n</g>\n<!-- C_1&#45;&gt;H_1 -->\n<g id=\"edge12\" class=\"edge\">\n<title>C_1&#45;&gt;H_1</title>\n<path fill=\"none\" stroke=\"black\" d=\"M135.7,-341.37C125.33,-332.59 113.4,-320.17 107.74,-306 95.87,-276.28 93.64,-190.72 107.74,-162 111.35,-154.64 117.27,-148.37 123.7,-143.21\"/>\n<polygon fill=\"black\" stroke=\"black\" points=\"125.88,-145.95 131.98,-137.29 121.81,-140.26 125.88,-145.95\"/>\n</g>\n<!-- C_1&#45;&gt;T_1 -->\n<g id=\"edge13\" class=\"edge\">\n<title>C_1&#45;&gt;T_1</title>\n<path fill=\"none\" stroke=\"black\" d=\"M135.7,-341.37C125.33,-332.59 113.4,-320.17 107.74,-306 86.96,-254 83.06,-104.27 107.74,-54 111.42,-46.52 117.48,-40.15 124.04,-34.94\"/>\n<polygon fill=\"black\" stroke=\"black\" points=\"126.33,-37.61 132.46,-28.97 122.28,-31.9 126.33,-37.61\"/>\n</g>\n<!-- C_2 -->\n<g id=\"node20\" class=\"node\">\n<title>C_2</title>\n<ellipse fill=\"none\" stroke=\"black\" cx=\"247.53\" cy=\"-409\" rx=\"27.9\" ry=\"18\"/>\n<text text-anchor=\"middle\" x=\"247.53\" y=\"-405.3\" font-family=\"Times,serif\" font-size=\"14.00\">C_2</text>\n</g>\n<!-- C_1&#45;&gt;C_2 -->\n<g id=\"edge24\" class=\"edge\">\n<title>C_1&#45;&gt;C_2</title>\n<path fill=\"none\" stroke=\"black\" d=\"M175.58,-366.98C187.98,-374.33 204.04,-383.84 217.75,-391.95\"/>\n<polygon fill=\"black\" stroke=\"black\" points=\"215.96,-394.97 226.35,-397.05 219.53,-388.94 215.96,-394.97\"/>\n</g>\n<!-- P_2 -->\n<g id=\"node15\" class=\"node\">\n<title>P_2</title>\n<ellipse fill=\"none\" stroke=\"black\" cx=\"247.53\" cy=\"-355\" rx=\"27.1\" ry=\"18\"/>\n<text text-anchor=\"middle\" x=\"247.53\" y=\"-351.3\" font-family=\"Times,serif\" font-size=\"14.00\">P_2</text>\n</g>\n<!-- A_2 -->\n<g id=\"node16\" class=\"node\">\n<title>A_2</title>\n<ellipse fill=\"none\" stroke=\"black\" cx=\"247.53\" cy=\"-139\" rx=\"27.1\" ry=\"18\"/>\n<text text-anchor=\"middle\" x=\"247.53\" y=\"-135.3\" font-family=\"Times,serif\" font-size=\"14.00\">A_2</text>\n</g>\n<!-- P_2&#45;&gt;A_2 -->\n<g id=\"edge15\" class=\"edge\">\n<title>P_2&#45;&gt;A_2</title>\n<path fill=\"none\" stroke=\"black\" d=\"M225.66,-344.03C216.22,-338.03 206.16,-329.64 200.94,-319 186.83,-290.28 186.83,-203.72 200.94,-175 204.61,-167.52 210.68,-161.15 217.23,-155.94\"/>\n<polygon fill=\"black\" stroke=\"black\" points=\"219.52,-158.61 225.66,-149.97 215.47,-152.9 219.52,-158.61\"/>\n</g>\n<!-- H_2 -->\n<g id=\"node17\" class=\"node\">\n<title>H_2</title>\n<ellipse fill=\"none\" stroke=\"black\" cx=\"247.53\" cy=\"-301\" rx=\"28.7\" ry=\"18\"/>\n<text text-anchor=\"middle\" x=\"247.53\" y=\"-297.3\" font-family=\"Times,serif\" font-size=\"14.00\">H_2</text>\n</g>\n<!-- P_2&#45;&gt;H_2 -->\n<g id=\"edge16\" class=\"edge\">\n<title>P_2&#45;&gt;H_2</title>\n<path fill=\"none\" stroke=\"black\" d=\"M247.53,-336.79C247.53,-334.31 247.53,-331.83 247.53,-329.34\"/>\n<polygon fill=\"black\" stroke=\"black\" points=\"251.03,-329.14 247.53,-319.14 244.03,-329.14 251.03,-329.14\"/>\n</g>\n<!-- T_2 -->\n<g id=\"node21\" class=\"node\">\n<title>T_2</title>\n<ellipse fill=\"none\" stroke=\"black\" cx=\"247.53\" cy=\"-85\" rx=\"27\" ry=\"18\"/>\n<text text-anchor=\"middle\" x=\"247.53\" y=\"-81.3\" font-family=\"Times,serif\" font-size=\"14.00\">T_2</text>\n</g>\n<!-- A_2&#45;&gt;T_2 -->\n<g id=\"edge21\" class=\"edge\">\n<title>A_2&#45;&gt;T_2</title>\n<path fill=\"none\" stroke=\"black\" d=\"M247.53,-120.79C247.53,-118.31 247.53,-115.83 247.53,-113.34\"/>\n<polygon fill=\"black\" stroke=\"black\" points=\"251.03,-113.14 247.53,-103.14 244.03,-113.14 251.03,-113.14\"/>\n</g>\n<!-- I_2 -->\n<g id=\"node18\" class=\"node\">\n<title>I_2</title>\n<ellipse fill=\"none\" stroke=\"black\" cx=\"247.53\" cy=\"-247\" rx=\"27\" ry=\"18\"/>\n<text text-anchor=\"middle\" x=\"247.53\" y=\"-243.3\" font-family=\"Times,serif\" font-size=\"14.00\">I_2</text>\n</g>\n<!-- I_2&#45;&gt;A_2 -->\n<g id=\"edge17\" class=\"edge\">\n<title>I_2&#45;&gt;A_2</title>\n<path fill=\"none\" stroke=\"black\" d=\"M225.66,-236.03C216.22,-230.03 206.16,-221.64 200.94,-211 193.88,-196.64 193.88,-189.36 200.94,-175 204.61,-167.52 210.68,-161.15 217.23,-155.94\"/>\n<polygon fill=\"black\" stroke=\"black\" points=\"219.52,-158.61 225.66,-149.97 215.47,-152.9 219.52,-158.61\"/>\n</g>\n<!-- I_2&#45;&gt;S_2 -->\n<g id=\"edge18\" class=\"edge\">\n<title>I_2&#45;&gt;S_2</title>\n<path fill=\"none\" stroke=\"black\" d=\"M247.53,-228.79C247.53,-226.31 247.53,-223.83 247.53,-221.34\"/>\n<polygon fill=\"black\" stroke=\"black\" points=\"251.03,-221.14 247.53,-211.14 244.03,-221.14 251.03,-221.14\"/>\n</g>\n<!-- C_2&#45;&gt;H_2 -->\n<g id=\"edge19\" class=\"edge\">\n<title>C_2&#45;&gt;H_2</title>\n<path fill=\"none\" stroke=\"black\" d=\"M225.17,-397.71C215.88,-391.74 206.07,-383.46 200.94,-373 193.88,-358.64 193.88,-351.36 200.94,-337 204.55,-329.64 210.47,-323.37 216.89,-318.21\"/>\n<polygon fill=\"black\" stroke=\"black\" points=\"219.08,-320.95 225.17,-312.29 215,-315.26 219.08,-320.95\"/>\n</g>\n<!-- C_2&#45;&gt;T_2 -->\n<g id=\"edge20\" class=\"edge\">\n<title>C_2&#45;&gt;T_2</title>\n<path fill=\"none\" stroke=\"black\" d=\"M225.17,-397.71C215.88,-391.74 206.07,-383.46 200.94,-373 188.59,-347.87 188.59,-146.13 200.94,-121 204.61,-113.52 210.68,-107.15 217.23,-101.94\"/>\n<polygon fill=\"black\" stroke=\"black\" points=\"219.52,-104.61 225.66,-95.97 215.47,-98.9 219.52,-104.61\"/>\n</g>\n</g>\n</svg>\n",
      "text/plain": [
       "<graphviz.sources.Source at 0x7f82e4598a30>"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dag_view"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Looks good! Now we need to generate the data. \n",
    "* The following cell uses YAWNING TITAN to set up 10 different environments, each with a random value for $P$ and $I$, and runs them for 50 timesteps. \n",
    "* As said in the paper, features of the enironment are then used to calculate values for each of the nodes in our DAG\n",
    "* The data is then in the `slice_data` dict, with keys for each node in the DAG"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "def create_env(use_same_net=False):\n",
    "    \"\"\"\n",
    "    Helper function to create an environment.\n",
    "\n",
    "    Args:\n",
    "        use_same_net: If true uses a saved network, otherwise creates a new network\n",
    "\n",
    "    Returns: The env\n",
    "\n",
    "    \"\"\"\n",
    "    settings_path = \"dcbo_config.yaml\"\n",
    "\n",
    "    if use_same_net:\n",
    "        matrix, node_positions = network_creator.load_network(\"dcbo_base_net.txt\")\n",
    "    else:\n",
    "        matrix, node_positions = network_creator.create_mesh(size=10)\n",
    "\n",
    "    network_interface = NetworkInterface(\n",
    "        matrix, node_positions, settings_path=settings_path\n",
    "    )\n",
    "\n",
    "    red = RedInterface(network_interface)\n",
    "    blue = BlueInterface(network_interface)\n",
    "\n",
    "    number_of_actions = blue.get_number_of_actions()\n",
    "\n",
    "    env = GenericNetworkEnv(\n",
    "        red,\n",
    "        blue,\n",
    "        network_interface,\n",
    "        number_of_actions,\n",
    "        show_metrics_every=10,\n",
    "        collect_additional_per_ts_data=True,\n",
    "    )\n",
    "\n",
    "    check_env(env, warn=True)\n",
    "\n",
    "    env.reset()\n",
    "\n",
    "    return env\n",
    "\n",
    "# Creates all 10 envs\n",
    "all_envs = [create_env() for i in range(10)]\n",
    "agent = DCBOAgent(all_envs[0].action_space, [0.5, 0.5])\n",
    "\n",
    "TIMESTEPS = 25\n",
    "COSTS = {\"restore_node\": 1, \"isolate\": 1, \"do_nothing\": 0, \"compromise\": 10}\n",
    "\n",
    "centralities = nx.degree_centrality(all_envs[0].network_interface.current_graph)\n",
    "slice_data = {k: [] for k in \"PICSHAT\"}\n",
    "\n",
    "for counter, current_env in enumerate(all_envs):\n",
    "    # Init data\n",
    "    slice_p, slice_i, slice_s, slice_h, slice_c, slice_a, slice_t = (list() for i in range(7))\n",
    "    \n",
    "    # Init probabilities\n",
    "    current_env.reset()\n",
    "    agent.reset()\n",
    "\n",
    "    w1, w2 = (\n",
    "        random.gauss(0.5, 0.167),\n",
    "        random.gauss(0.5, 0.167),\n",
    "    )\n",
    "    agent.update_probabilities([w1, w2])\n",
    "\n",
    "    for i in range(TIMESTEPS):\n",
    "        done = False\n",
    "        current_step = 0\n",
    "\n",
    "        action = agent.predict(\"\", \"\", \"\", current_env)\n",
    "        env_observation, reward, done, notes = current_env.step(action)\n",
    "\n",
    "        c_cost = (COSTS[\"compromise\"] * sum(notes[\"end_state\"].values())) ** 1.5\n",
    "        action_cost = COSTS.get(notes[\"blue_action\"], 0)\n",
    "        surface = sum(\n",
    "            map(\n",
    "                lambda x: not x[0] and not x[1],\n",
    "                zip(\n",
    "                    notes[\"end_state\"].values(),\n",
    "                    current_env.network_interface.get_all_isolation().values(),\n",
    "                ),\n",
    "            )\n",
    "        )\n",
    "        # Append P, I and S\n",
    "        slice_p.append(agent.probabilities[0])\n",
    "        slice_i.append(agent.probabilities[1])\n",
    "        slice_s.append(surface) \n",
    "        \n",
    "        # Calculate and append H\n",
    "        h = 0\n",
    "        node_states = current_env.network_interface.get_all_node_compromised_states()\n",
    "        node_iso = current_env.network_interface.get_all_isolation()\n",
    "        comp_nodes = [k for k in node_states if node_states[k] == 1]\n",
    "        isolated_nodes = [k for k in node_iso if node_iso[k]]\n",
    "        for node in comp_nodes:\n",
    "            conn_nodes = current_env.network_interface.get_current_connected_nodes(node)\n",
    "            for c in conn_nodes:\n",
    "                if c in isolated_nodes:\n",
    "                    continue\n",
    "                h += current_env.network_interface.get_single_node_vulnerability(c)\n",
    "        slice_h.append(h)\n",
    "        # Append C, A and T\n",
    "        slice_c.append(c_cost)\n",
    "        slice_a.append(action_cost)\n",
    "        slice_t.append(action_cost + c_cost)\n",
    "\n",
    "    slice_data[\"P\"].append(np.asarray(slice_p))\n",
    "    slice_data[\"I\"].append(np.asarray(slice_i))\n",
    "    slice_data[\"S\"].append(np.asarray(slice_s))\n",
    "    slice_data[\"H\"].append(np.asarray(slice_h))\n",
    "    slice_data[\"C\"].append(np.asarray(slice_c))\n",
    "    slice_data[\"A\"].append(np.asarray(slice_a))\n",
    "    slice_data[\"T\"].append(np.asarray(slice_t))\n",
    "\n",
    "for k in slice_data.keys():\n",
    "    slice_data[k] = np.asarray(slice_data[k])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Variables: dict_keys(['P', 'I', 'C', 'S', 'H', 'A', 'T'])\n",
      "Each variable has shape: (10, 25)\n"
     ]
    }
   ],
   "source": [
    "print(f\"Variables: {slice_data.keys()}\")\n",
    "print(f\"Each variable has shape: {slice_data['P'].shape}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Great, now let's use our full dataset to build our model\n",
    "* Since we have access to all the data, we can model the functional relationships between variables\n",
    "* We can then use these models to calculate the ground truch best value for each of our manipulative variables ($P$ and $I$) in the final 3 timesteps\n",
    "* We define a helper function for fitting GPs, then use this to fit a GP to each of the within-timeslice functions in our DAG"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "def fit_gp(\n",
    "    x,\n",
    "    y,\n",
    "    lengthscale=1.0,\n",
    "    variance=1.0,\n",
    "    noise_var=1.0,\n",
    "    ard=False,\n",
    "    n_restart=10,\n",
    "    seed: int = 0,\n",
    "):\n",
    "    \"\"\"\n",
    "    Helper function to fit a GP\n",
    "    \"\"\"\n",
    "    kernel = RBF(x.shape[1], ARD=ard, lengthscale=lengthscale, variance=variance)\n",
    "    model = GPRegression(X=x, Y=y, kernel=kernel, noise_var=noise_var)\n",
    "    model.optimize_restarts(n_restart, verbose=False, robust=True)\n",
    "    return model\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "time_vector = np.tile(\n",
    "    np.linspace(0, slice_data[\"P\"].shape[1] - 1, slice_data[\"P\"].shape[1]),\n",
    "    10\n",
    ")[:, np.newaxis]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " /home/sam/miniconda3/envs/bt/lib/python3.9/site-packages/GPy/kern/src/stationary.py:168: RuntimeWarning:overflow encountered in true_divide\n",
      " /home/sam/miniconda3/envs/bt/lib/python3.9/site-packages/GPy/kern/src/rbf.py:52: RuntimeWarning:overflow encountered in square\n",
      " /home/sam/miniconda3/envs/bt/lib/python3.9/site-packages/GPy/kern/src/rbf.py:76: RuntimeWarning:invalid value encountered in multiply\n",
      " /home/sam/miniconda3/envs/bt/lib/python3.9/site-packages/paramz/transformations.py:113: RuntimeWarning:invalid value encountered in expm1\n"
     ]
    }
   ],
   "source": [
    "# Fit our GPs for t=0 (so no inter-timeslice dependencies)\n",
    "model_P = fit_gp(time_vector, np.hstack(slice_data[\"P\"])[:, np.newaxis])\n",
    "model_I = fit_gp(time_vector, np.hstack(slice_data[\"I\"])[:, np.newaxis])\n",
    "model_S = fit_gp(\n",
    "    np.hstack(slice_data[\"I\"])[:, np.newaxis], np.hstack(slice_data[\"S\"])[:, np.newaxis]\n",
    ")\n",
    "model_C = fit_gp(time_vector, np.hstack(slice_data[\"C\"])[:, np.newaxis])\n",
    "model_H = fit_gp(\n",
    "    np.hstack(\n",
    "        (\n",
    "            np.hstack(slice_data[\"P\"])[:, np.newaxis],\n",
    "            np.hstack(slice_data[\"C\"])[:, np.newaxis],\n",
    "        )\n",
    "    ), \n",
    "    np.hstack(slice_data[\"H\"])[:, np.newaxis],\n",
    ")\n",
    "model_A = fit_gp(\n",
    "    np.hstack(\n",
    "        (\n",
    "            np.hstack(slice_data[\"P\"])[:, np.newaxis],\n",
    "            np.hstack(slice_data[\"I\"])[:, np.newaxis],\n",
    "        )\n",
    "    ),\n",
    "    np.hstack(slice_data[\"A\"])[:, np.newaxis],\n",
    ")\n",
    "model_T = fit_gp(\n",
    "    np.hstack(\n",
    "        (\n",
    "            np.hstack(slice_data[\"C\"])[:, np.newaxis],\n",
    "            np.hstack(slice_data[\"A\"])[:, np.newaxis],\n",
    "        )\n",
    "    ),\n",
    "    np.hstack(slice_data[\"T\"])[:, np.newaxis],\n",
    ")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "functions_0 = {\n",
    "    \"P\": model_P,\n",
    "    \"I\": model_I,\n",
    "    \"S\": model_S,\n",
    "    \"H\": model_H,\n",
    "    \"C\": model_C,\n",
    "    \"A\": model_A,\n",
    "    \"T\": model_T,\n",
    "}"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Let's visualise our GPs to check they look sensible:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " /home/sam/miniconda3/envs/bt/lib/python3.9/site-packages/GPy/plotting/matplot_dep/plot_definitions.py:184: UserWarning:The following kwargs were not used by contour: 'label', 'linewidth'\n",
      " /home/sam/miniconda3/envs/bt/lib/python3.9/site-packages/GPy/plotting/matplot_dep/plot_definitions.py:184: UserWarning:The following kwargs were not used by contour: 'label', 'linewidth'\n",
      " /home/sam/miniconda3/envs/bt/lib/python3.9/site-packages/GPy/likelihoods/gaussian.py:119: RuntimeWarning:invalid value encountered in sqrt\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "for k, m in functions_0.items():\n",
    "    m.plot()\n",
    "    plt.suptitle(k)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### And let's also plot some of the variables from one of our environments, to see how it looks:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x7f82c51eea90>"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "for var in \"SCHAT\":\n",
    "    plt.plot(np.linspace(0, 24, 25), slice_data[var][1], label=var)\n",
    "plt.legend()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Okay, now we need to model the functions for $t > 0$, again using GPs.\n",
    "* First, we rearrange the data so that it's easy to fit with our helper\n",
    "  * Note that when we construct our arrays (e.g. `S_AR` in the below cell), we are constructing them so they map to the relationships in the DAG (e.g. `S_AR` contains data from $C_{t-1}$ and $I_{t}$, where $t$ is the current timeslice)\n",
    "* Then, we fit"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Getting slice_data in the right form, based on the DAG\n",
    "S_AR, H_AR, C_AR, A_AR, T_AR = [], [], [], [], []\n",
    "\n",
    "for i in range(slice_data[\"P\"].shape[0]):\n",
    "    S_AR.append(\n",
    "        np.transpose(np.vstack((slice_data[\"C\"][i][:-1],slice_data[\"I\"][i][1:],slice_data[\"S\"][i][1:])))\n",
    "    )\n",
    "    C_AR.append(\n",
    "        np.transpose(np.vstack((slice_data[\"H\"][i][:-1],slice_data[\"C\"][i][1:])))\n",
    "    )\n",
    "    H_AR.append(\n",
    "        np.transpose(np.vstack((slice_data[\"P\"][i][1:],slice_data[\"C\"][i][1:],slice_data[\"H\"][i][1:])))\n",
    "    )\n",
    "    A_AR.append(\n",
    "        np.transpose(np.vstack((slice_data[\"P\"][i][1:],slice_data[\"I\"][i][1:],slice_data[\"A\"][i][1:])))\n",
    "    )\n",
    "    T_AR.append(\n",
    "        np.transpose(np.vstack((slice_data[\"C\"][i][1:],slice_data[\"A\"][i][1:],slice_data[\"T\"][i][1:])))\n",
    "    )\n",
    "\n",
    "S_AR = np.vstack(S_AR)\n",
    "H_AR = np.vstack(H_AR)\n",
    "C_AR = np.vstack(C_AR)\n",
    "A_AR = np.vstack(A_AR)\n",
    "T_AR = np.vstack(T_AR)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " /home/sam/miniconda3/envs/bt/lib/python3.9/site-packages/GPy/kern/src/rbf.py:52: RuntimeWarning:overflow encountered in square\n",
      " /home/sam/miniconda3/envs/bt/lib/python3.9/site-packages/GPy/kern/src/stationary.py:168: RuntimeWarning:overflow encountered in true_divide\n",
      " /home/sam/miniconda3/envs/bt/lib/python3.9/site-packages/GPy/kern/src/rbf.py:76: RuntimeWarning:invalid value encountered in multiply\n",
      " /home/sam/miniconda3/envs/bt/lib/python3.9/site-packages/paramz/transformations.py:113: RuntimeWarning:invalid value encountered in expm1\n"
     ]
    }
   ],
   "source": [
    "model_S_t = fit_gp(S_AR[:, :-1], S_AR[:, -1][:, np.newaxis])\n",
    "model_H_t = fit_gp(H_AR[:, :-1], H_AR[:, -1][:, np.newaxis])\n",
    "model_C_t = fit_gp(C_AR[:, :-1], C_AR[:, -1][:, np.newaxis])\n",
    "model_A_t = fit_gp(A_AR[:, :-1], A_AR[:, -1][:, np.newaxis])\n",
    "model_T_t = fit_gp(T_AR[:, :-1], T_AR[:, -1][:, np.newaxis])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "functions_t = {\n",
    "    \"P\": model_P,\n",
    "    \"I\": model_I,\n",
    "    \"S\": model_S_t,\n",
    "    \"H\": model_H_t,\n",
    "    \"C\": model_C_t,\n",
    "    \"A\": model_A_t,\n",
    "    \"T\": model_T_t,\n",
    "}\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Again, let's plot some of these functions to check they look sensible"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " /tmp/ipykernel_33740/248468357.py:10: MatplotlibDeprecationWarning:Calling gca() with keyword arguments was deprecated in Matplotlib 3.4. Starting two minor releases later, gca() will take no keyword arguments. The gca() function should only be used to get the current axes, or if no axes exist, create new axes with default keyword arguments. To create a new axes with non-default arguments, use plt.axes() or plt.subplot().\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "C, P = np.meshgrid(\n",
    "    np.linspace(-0.0, 0.1, 100)[:, np.newaxis],\n",
    "    np.linspace(-0.0, 1.0, 100)[:, np.newaxis],\n",
    ")\n",
    "H_vals = model_H_t.predict(\n",
    "    np.transpose(np.vstack((np.hstack(P), np.hstack(C))))\n",
    ")[0]\n",
    "\n",
    "fig = plt.figure()\n",
    "ax = fig.gca(projection=\"3d\")\n",
    "ax.plot_surface(P, C, H_vals.reshape(100, 100))\n",
    "plt.xlabel(\"$P_t$\")\n",
    "plt.ylabel(\"$C_t$\")\n",
    "plt.suptitle(\"Plot of $H_t$ vs. $P_t$ and $C_{t}}$\")\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " /tmp/ipykernel_33740/3295890710.py:8: MatplotlibDeprecationWarning:Calling gca() with keyword arguments was deprecated in Matplotlib 3.4. Starting two minor releases later, gca() will take no keyword arguments. The gca() function should only be used to get the current axes, or if no axes exist, create new axes with default keyword arguments. To create a new axes with non-default arguments, use plt.axes() or plt.subplot().\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "I, C = np.meshgrid(\n",
    "    np.linspace(-1.0, 100, 100)[:, np.newaxis],\n",
    "    np.linspace(-1.0, 100, 100)[:, np.newaxis],\n",
    ")\n",
    "S_vals = model_S_t.predict(np.transpose(np.vstack((np.hstack(C), np.hstack(I)))))[0]\n",
    "\n",
    "fig = plt.figure()\n",
    "ax = fig.gca(projection=\"3d\")\n",
    "ax.plot_surface(C, I, S_vals.reshape(100, 100))\n",
    "plt.xlabel(\"C\")\n",
    "plt.ylabel(\"I\")\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Great, so now we have models for all of the functions within our DAG, we can use them to build up our Structural Equation Model\n",
    "* The `static` method of our SEM uses our $t = 0$ models to define the relationships between variables in the DAG in a static setting (e.g. just looking at one time slice)\n",
    "* The `dynamic` method of our SEM uses our $t > 0$ models to define the relationships between variables in the DAG in a dynamic setting (e.g. taking account of temporal relationships)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "class PICSHAT_SEM:\n",
    "    def __init__(self, functions_0, functions_t):\n",
    "        self.functions_0 = functions_0\n",
    "        self.functions_t = functions_t\n",
    "\n",
    "    def static(self):\n",
    "\n",
    "        P = (\n",
    "            lambda noise, time, sample: noise\n",
    "            + self.functions_0[\"P\"].predict(time * np.ones((1, 1)))[0]\n",
    "        )\n",
    "\n",
    "        I = (\n",
    "            lambda noise, time, sample: noise\n",
    "            + self.functions_0[\"I\"].predict(time * np.ones((1, 1)))[0]\n",
    "        )\n",
    "\n",
    "        S = (\n",
    "            lambda noise, time, sample: noise\n",
    "            + self.functions_0[\"S\"].predict(\n",
    "                np.transpose(np.vstack((sample[\"I\"][time] * np.ones((1, 1)))))\n",
    "            )[0]\n",
    "        )\n",
    "\n",
    "        C = (\n",
    "            lambda noise, time, sample: noise\n",
    "            + self.functions_0[\"C\"].predict(time * np.ones((1, 1)))[0]\n",
    "        )\n",
    "\n",
    "        H = (\n",
    "            lambda noise, time, sample: noise\n",
    "            + self.functions_0[\"H\"].predict(\n",
    "                np.transpose(\n",
    "                    np.vstack(\n",
    "                        (\n",
    "                            sample[\"P\"][time] * np.ones((1, 1)),\n",
    "                            sample[\"C\"][time] * np.ones((1, 1)),\n",
    "                        )\n",
    "                    )\n",
    "                )\n",
    "            )[0]\n",
    "        )\n",
    "\n",
    "        A = (\n",
    "            lambda noise, time, sample: noise\n",
    "            + self.functions_0[\"A\"].predict(\n",
    "                np.transpose(\n",
    "                    np.vstack(\n",
    "                        (\n",
    "                            sample[\"P\"][time] * np.ones((1, 1)),\n",
    "                            sample[\"I\"][time] * np.ones((1, 1)),\n",
    "                        )\n",
    "                    )\n",
    "                )\n",
    "            )[0]\n",
    "        )\n",
    "\n",
    "        T = (\n",
    "            lambda noise, time, sample: np.abs(\n",
    "                noise\n",
    "                + self.functions_0[\"T\"].predict(\n",
    "                    np.transpose(\n",
    "                        np.vstack(\n",
    "                            (\n",
    "                                sample[\"C\"][time] * np.ones((1, 1)),\n",
    "                                sample[\"A\"][time] * np.ones((1, 1)),\n",
    "                            )\n",
    "                        )\n",
    "                    )\n",
    "                )[0]\n",
    "            )\n",
    "        )\n",
    "\n",
    "        return OrderedDict(\n",
    "            [(\"P\", P), (\"I\", I), (\"S\", S), (\"H\", H), (\"C\", C), (\"A\", A), (\"T\", T)]\n",
    "        )\n",
    "\n",
    "    def dynamic(self):\n",
    "\n",
    "        P = (\n",
    "            lambda noise, time, sample: noise\n",
    "            + self.functions_t[\"P\"].predict(time * np.ones((1, 1)))[0]\n",
    "        )\n",
    "\n",
    "        I = (\n",
    "            lambda noise, time, sample: noise\n",
    "            + self.functions_t[\"I\"].predict(time * np.ones((1, 1)))[0]\n",
    "        )\n",
    "\n",
    "        S = (\n",
    "            lambda noise, time, sample: noise\n",
    "            + self.functions_t[\"S\"].predict(\n",
    "                np.transpose(\n",
    "                    np.vstack(\n",
    "                        (\n",
    "                            sample[\"C\"][time - 1] * np.ones((1, 1)),\n",
    "                            sample[\"I\"][time] * np.ones((1, 1)),\n",
    "                        )\n",
    "                    )\n",
    "                )\n",
    "            )[0]\n",
    "        )\n",
    "\n",
    "        C = (\n",
    "            lambda noise, time, sample: noise\n",
    "            + self.functions_t[\"C\"].predict(\n",
    "                np.transpose(\n",
    "                    np.vstack(\n",
    "                        (\n",
    "                            sample[\"H\"][time - 1] * np.ones((1,1)),\n",
    "                        )\n",
    "                    )\n",
    "                )\n",
    "            )[0]\n",
    "        )\n",
    "\n",
    "        H = (\n",
    "            lambda noise, time, sample: noise\n",
    "            + self.functions_t[\"H\"].predict(\n",
    "                np.transpose(\n",
    "                    np.vstack(\n",
    "                        (\n",
    "                            sample[\"P\"][time] * np.ones((1, 1)),\n",
    "                            sample[\"C\"][time] * np.ones((1, 1)),\n",
    "                        )\n",
    "                    )\n",
    "                )\n",
    "            )[0]\n",
    "        )\n",
    "\n",
    "        A = (\n",
    "            lambda noise, time, sample: noise\n",
    "            + self.functions_t[\"A\"].predict(\n",
    "                np.transpose(\n",
    "                    np.vstack(\n",
    "                        (\n",
    "                            sample[\"P\"][time] * np.ones((1, 1)),\n",
    "                            sample[\"I\"][time] * np.ones((1, 1)),\n",
    "                        )\n",
    "                    )\n",
    "                )\n",
    "            )[0]\n",
    "        )\n",
    "\n",
    "        T = lambda noise, time, sample: np.abs(\n",
    "            noise\n",
    "            + self.functions_t[\"T\"].predict(\n",
    "                np.transpose(\n",
    "                    np.vstack(\n",
    "                        (\n",
    "                            sample[\"C\"][time] * np.ones((1, 1)),\n",
    "                            sample[\"A\"][time] * np.ones((1, 1)),\n",
    "                        )\n",
    "                    )\n",
    "                )\n",
    "            )[0]\n",
    "        )\n",
    "\n",
    "        return OrderedDict(\n",
    "            [(\"P\", P), (\"I\", I), (\"S\", S), (\"H\", H), (\"C\", C), (\"A\", A), (\"T\", T)]\n",
    "        )\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Here, we set up some parameters for the optimisation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "intervention_domain = {\"P\": [0.0, 1.0], \"I\": [0.0, 1.0]}\n",
    "exploration_sets = list(powerset([\"P\", \"I\"]))\n",
    "n_to_compute = 50\n",
    "interventional_grids = get_interventional_grids(\n",
    "    exploration_sets, intervention_domain, size_intervention_grid=n_to_compute\n",
    ")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "SEM = PICSHAT_SEM(functions_0, functions_t)\n",
    "initial_structural_equation_model = SEM.static()\n",
    "structural_equation_model = SEM.dynamic()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### In the following cell, we use our SEM to calculate the ground truth optimal value for our interventional variables, $P$ and $I$\n",
    "* N.b. - this can take a while to run, as it is searching the whole intervention space for the optimal value"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "(\n",
    "    best_s_values,\n",
    "    best_s_sequence,\n",
    "    best_objective_values,\n",
    "    y_stars_all,\n",
    "    optimal_interventions,\n",
    "    all_CE,\n",
    ") = optimal_sequence_of_interventions(\n",
    "    exploration_sets,\n",
    "    interventional_grids,\n",
    "    initial_structural_equation_model,\n",
    "    structural_equation_model,\n",
    "    dag,\n",
    "    T=T,\n",
    "    model_variables=list(\"PISHCAT\"),\n",
    "    target_variable=\"T\",\n",
    "    task=\"min\",\n",
    ")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### So now we have our ground truth optimal interventions:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best intervention set: [('P', 'I'), ('P', 'I'), ('P', 'I')]\n",
      "Best intervention values: [array([0.36734694, 0.81632653]), array([1., 0.]), array([1., 0.])]\n",
      "Best objective values: [1.1532104545822945, 12.88861608505249, 12.88861608505249]\n"
     ]
    }
   ],
   "source": [
    "print(\"Best intervention set:\", best_s_sequence)\n",
    "print(\"Best intervention values:\", best_s_values)\n",
    "print(\"Best objective values:\", best_objective_values)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Now we're going to actually run the experiment.\n",
    "* We run each optimisation algorithm, giving each one 50 samples to converge on it's optimal objective value for each of the final 3 timesteps\n",
    "* We run each algorithm 3 times, so we can plot uncertainty bounds and averge performance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "subset_obs_data = deepcopy(slice_data)\n",
    "for var in subset_obs_data.keys():\n",
    "    subset_obs_data[var] = subset_obs_data[var][:, -3:]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Experiment count:   0%|          | 0/3 [00:00<?, ?it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "\t>>>DCBO\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " /home/sam/miniconda3/envs/bt/lib/python3.9/site-packages/GPy/kern/src/rbf.py:52: RuntimeWarning:overflow encountered in square\n",
      " /home/sam/miniconda3/envs/bt/lib/python3.9/site-packages/GPy/kern/src/stationary.py:168: RuntimeWarning:overflow encountered in true_divide\n",
      " /home/sam/miniconda3/envs/bt/lib/python3.9/site-packages/GPy/kern/src/rbf.py:76: RuntimeWarning:invalid value encountered in multiply\n",
      " /home/sam/miniconda3/envs/bt/lib/python3.9/site-packages/paramz/transformations.py:113: RuntimeWarning:invalid value encountered in expm1\n",
      " ../DCBO/src/bayes_opt/causal_kernels.py:79: RuntimeWarning:invalid value encountered in sqrt\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      " /home/sam/miniconda3/envs/bt/lib/python3.9/site-packages/paramz/transformations.py:119: RuntimeWarning:invalid value encountered in expm1\n",
      " /home/sam/miniconda3/envs/bt/lib/python3.9/site-packages/paramz/transformations.py:121: RuntimeWarning:invalid value encountered in expm1\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "Time index: 100%|██████████| 3/3 [05:24<00:00, 108.25s/it]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "\t>>>CBO\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " /home/sam/miniconda3/envs/bt/lib/python3.9/site-packages/GPy/kern/src/rbf.py:52: RuntimeWarning:overflow encountered in square\n",
      " /home/sam/miniconda3/envs/bt/lib/python3.9/site-packages/GPy/kern/src/stationary.py:168: RuntimeWarning:overflow encountered in true_divide\n",
      " /home/sam/miniconda3/envs/bt/lib/python3.9/site-packages/GPy/kern/src/rbf.py:76: RuntimeWarning:invalid value encountered in multiply\n",
      " /home/sam/miniconda3/envs/bt/lib/python3.9/site-packages/paramz/transformations.py:113: RuntimeWarning:invalid value encountered in expm1\n",
      " ../DCBO/src/bayes_opt/causal_kernels.py:79: RuntimeWarning:invalid value encountered in sqrt\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      " /home/sam/miniconda3/envs/bt/lib/python3.9/site-packages/paramz/transformations.py:119: RuntimeWarning:invalid value encountered in expm1\n",
      " /home/sam/miniconda3/envs/bt/lib/python3.9/site-packages/paramz/transformations.py:121: RuntimeWarning:invalid value encountered in expm1\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "Time index: 100%|██████████| 3/3 [06:29<00:00, 129.99s/it]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "\t>>>BO\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "Time index: 100%|██████████| 3/3 [00:05<00:00,  1.71s/it]\n",
      "Experiment count:  33%|███▎      | 1/3 [12:38<25:16, 758.17s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "\t>>>DCBO\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " /home/sam/miniconda3/envs/bt/lib/python3.9/site-packages/GPy/kern/src/rbf.py:52: RuntimeWarning:overflow encountered in square\n",
      " /home/sam/miniconda3/envs/bt/lib/python3.9/site-packages/GPy/kern/src/stationary.py:168: RuntimeWarning:overflow encountered in true_divide\n",
      " /home/sam/miniconda3/envs/bt/lib/python3.9/site-packages/GPy/kern/src/rbf.py:76: RuntimeWarning:invalid value encountered in multiply\n",
      " /home/sam/miniconda3/envs/bt/lib/python3.9/site-packages/paramz/transformations.py:113: RuntimeWarning:invalid value encountered in expm1\n",
      " ../DCBO/src/bayes_opt/causal_kernels.py:79: RuntimeWarning:invalid value encountered in sqrt\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      " /home/sam/miniconda3/envs/bt/lib/python3.9/site-packages/paramz/transformations.py:119: RuntimeWarning:invalid value encountered in expm1\n",
      " /home/sam/miniconda3/envs/bt/lib/python3.9/site-packages/paramz/transformations.py:121: RuntimeWarning:invalid value encountered in expm1\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "Time index: 100%|██████████| 3/3 [05:17<00:00, 105.97s/it]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "\t>>>CBO\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " /home/sam/miniconda3/envs/bt/lib/python3.9/site-packages/GPy/kern/src/rbf.py:52: RuntimeWarning:overflow encountered in square\n",
      " /home/sam/miniconda3/envs/bt/lib/python3.9/site-packages/GPy/kern/src/stationary.py:168: RuntimeWarning:overflow encountered in true_divide\n",
      " /home/sam/miniconda3/envs/bt/lib/python3.9/site-packages/GPy/kern/src/rbf.py:76: RuntimeWarning:invalid value encountered in multiply\n",
      " /home/sam/miniconda3/envs/bt/lib/python3.9/site-packages/paramz/transformations.py:113: RuntimeWarning:invalid value encountered in expm1\n",
      " ../DCBO/src/bayes_opt/causal_kernels.py:79: RuntimeWarning:invalid value encountered in sqrt\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      " /home/sam/miniconda3/envs/bt/lib/python3.9/site-packages/paramz/transformations.py:119: RuntimeWarning:invalid value encountered in expm1\n",
      " /home/sam/miniconda3/envs/bt/lib/python3.9/site-packages/paramz/transformations.py:121: RuntimeWarning:invalid value encountered in expm1\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "Time index: 100%|██████████| 3/3 [06:23<00:00, 127.99s/it]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "\t>>>BO\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "Time index: 100%|██████████| 3/3 [00:05<00:00,  1.74s/it]\n",
      "Experiment count:  67%|██████▋   | 2/3 [25:03<12:30, 750.45s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "\t>>>DCBO\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " /home/sam/miniconda3/envs/bt/lib/python3.9/site-packages/GPy/kern/src/rbf.py:52: RuntimeWarning:overflow encountered in square\n",
      " /home/sam/miniconda3/envs/bt/lib/python3.9/site-packages/GPy/kern/src/stationary.py:168: RuntimeWarning:overflow encountered in true_divide\n",
      " /home/sam/miniconda3/envs/bt/lib/python3.9/site-packages/GPy/kern/src/rbf.py:76: RuntimeWarning:invalid value encountered in multiply\n",
      " /home/sam/miniconda3/envs/bt/lib/python3.9/site-packages/paramz/transformations.py:113: RuntimeWarning:invalid value encountered in expm1\n",
      " ../DCBO/src/bayes_opt/causal_kernels.py:79: RuntimeWarning:invalid value encountered in sqrt\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      " /home/sam/miniconda3/envs/bt/lib/python3.9/site-packages/paramz/transformations.py:119: RuntimeWarning:invalid value encountered in expm1\n",
      " /home/sam/miniconda3/envs/bt/lib/python3.9/site-packages/paramz/transformations.py:121: RuntimeWarning:invalid value encountered in expm1\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "Time index: 100%|██████████| 3/3 [05:39<00:00, 113.32s/it]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "\t>>>CBO\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " /home/sam/miniconda3/envs/bt/lib/python3.9/site-packages/GPy/kern/src/rbf.py:52: RuntimeWarning:overflow encountered in square\n",
      " /home/sam/miniconda3/envs/bt/lib/python3.9/site-packages/GPy/kern/src/stationary.py:168: RuntimeWarning:overflow encountered in true_divide\n",
      " /home/sam/miniconda3/envs/bt/lib/python3.9/site-packages/GPy/kern/src/rbf.py:76: RuntimeWarning:invalid value encountered in multiply\n",
      " /home/sam/miniconda3/envs/bt/lib/python3.9/site-packages/paramz/transformations.py:113: RuntimeWarning:invalid value encountered in expm1\n",
      " ../DCBO/src/bayes_opt/causal_kernels.py:79: RuntimeWarning:invalid value encountered in sqrt\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      " /home/sam/miniconda3/envs/bt/lib/python3.9/site-packages/paramz/transformations.py:119: RuntimeWarning:invalid value encountered in expm1\n",
      " /home/sam/miniconda3/envs/bt/lib/python3.9/site-packages/paramz/transformations.py:121: RuntimeWarning:invalid value encountered in expm1\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "Time index: 100%|██████████| 3/3 [06:34<00:00, 131.40s/it]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "\t>>>BO\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "reconstraining parameters GP_regression.rbf.variance\n",
      "Time index: 100%|██████████| 3/3 [00:05<00:00,  1.68s/it]\n",
      "Experiment count: 100%|██████████| 3/3 [38:02<00:00, 760.77s/it]\n"
     ]
    }
   ],
   "source": [
    "# See method for argument details\n",
    "R, N = 3, 50\n",
    "results = run_methods_replicates(\n",
    "    G=dag,\n",
    "    sem=PICSHAT_SEM,\n",
    "    make_sem_estimator=build_sem_hat,\n",
    "    base_target_variable='T',\n",
    "    intervention_domain = intervention_domain,\n",
    "    methods_list = ['DCBO', 'CBO', 'BO'],\n",
    "    obs_samples = subset_obs_data,\n",
    "    exploration_sets = exploration_sets,\n",
    "    total_timesteps = T,\n",
    "    number_of_trials = N,\n",
    "    reps = R, # Number of replicates (how many times we run each method)\n",
    "    n_restart = 1,\n",
    "    save_data = False,\n",
    "    n_obs = 5, # The method samples 5 time-series for each replicate\n",
    "    num_anchor_points = 100,\n",
    "    sample_anchor_points = True,\n",
    "    controlled_experiment=True,\n",
    "    args_sem=[functions_0, functions_t],\n",
    "    manipulative_variables=[\"P\", \"I\"]\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Now, we collect our data from the experiment, and plot it"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Since we didn't save the results we cannot use the pickled file so we have to convert results to the correct format\n",
    "data = get_relevant_results(results=results,replicates=R)\n",
    "exp_optimal_outcome_values_during_trials, exp_per_trial_cost = elaborate(number_of_interventions=None, \n",
    "                                                                         n_replicates=R, \n",
    "                                                                         data=data, \n",
    "                                                                         best_objective_values=best_objective_values, \n",
    "                                                                         T=T)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "plot_params = {\n",
    "    \"linewidth\": 3,\n",
    "    \"linewidth_opt\": 4,\n",
    "    \"alpha\": 0.2,\n",
    "    \"xlim_max\": N,\n",
    "    \"ncols\": 5,\n",
    "    \"loc_legend\": \"lower right\",\n",
    "    \"size_ticks\": 20,\n",
    "    \"size_labels\": 20,\n",
    "    \"xlabel\": r'$\\texttt{cost}(\\mathbf{X}_{s,t}, \\mathbf{x}_{s,t})$',\n",
    "    \"labels\": {'DCBO': 'DCBO', 'CBO': 'CBO', 'ABO': 'ABO', 'BO': 'BO', 'True': r'$\\mathbb{E} \\left [Y_t \\mid \\textrm{do}(\\mathbf{X}_{s,t}^\\star = \\mathbf{x}_{s,t}^\\star) \\right]$'},\n",
    "    \"colors\": {'DCBO': 'blue', 'CBO': 'green', 'ABO': 'orange', 'BO': 'red', 'True': 'black'},\n",
    "    \"line_styles\": {'DCBO': '-', 'CBO': '--', 'ABO': 'dashdot', 'BO': '-', 'True': ':'},\n",
    "    \"width\":10\n",
    "}\n",
    "rc('text', usetex=True)\n",
    "rc('text.latex', preamble=r'\\usepackage{amssymb}')\n",
    "rc('font', family='serif')\n",
    "rc('font', size=20)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "class PlainOffsetScalarFormatter(ScalarFormatter):\n",
    "    def get_offset(self):\n",
    "        if len(self.locs) == 0:\n",
    "            return ''\n",
    "        if self.orderOfMagnitude:\n",
    "            pass\n",
    "            # print(\"Your plot will likely be labelled incorrectly\")\n",
    "        return \"$\\\\times 10^{-5}+$\" + str(round(self.offset, 5))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_lim_list = {0: (1, 1.5), 1: (12.8, 13.8), 2: (12.8, 13.8)}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABFAAAACoCAYAAADHCRpbAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjQuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8rg+JYAAAACXBIWXMAAAsTAAALEwEAmpwYAABy2UlEQVR4nO3deXgb13kv/u9gJbGS1L5AsqxdIL1Iom1CduxEskUqcRIzjqmkTSPWotTFFtOWvPemtXhv6bT31mTSUk37iwglUuo2EZxGSdNGhGzZtRwLlK3FlklQ+8bRQkncAHAngPn9cTgABgsJrgDJ9/M8eMjZzwDg4cw757yHEwRBACGEEEIIIYQQQgiJSZboAhBCCCGEEEIIIYQkOwqgEEIIIYQQQgghhAyBAiiEEEIIIYQQQgghQ6AACiGEEEIIIYQQQsgQKIBCCCGEEEIIIYQQMgRFogswEVauXJnoIhBCEuzChQsJPT7VQ4QQqocIIYmW6HqIkMluWgRQgPgqi6NnLuORpfMx06iZgBIRQiZKstw00EULIdMX1UOEkERLlnqIkMmMuvAM+PTyHbxf34BPr9xOdFEIIYQQQgghhBCSZCiAAqCptQOHT9dh2bx5cDbegt8vJLpIhBBCCCGEEEIISSLTPoDS0+eF7YPTmG0wQCGTodfbj6a2jkQXixBCCCGEEEIIIUlkWgdQ/H4Bv3LUQSbjkKJSAgBSVSrUXbuT4JIRQgghhBBCCCEkmUzrAErtuUbwzS1I12oD84waDRr429SNhxBCCCGEEEIIIQHTZhQec+mfSqa1ijQ8tvBRPLbyAcl8Gcehz+vF7RYPFs4yTGAJCUl+TqcTxcXFOHr0aEKOXV9fD5PJBLfbDYPBAIvFMuHlIIQkViLrIdGuXbuwc+dOmM3mhJWBEJI4iayH7HY7XC4XGhsb0dDQgIKCAuTm5k54OQiZrqZNAKWx47JkWgDQ2dWPJ1YtiVhXq1bj0yu3KIBCSJjDhw/DZDJN+HF5nsfevXuxZ8+ewLxdu3bBaDTSDQwh00wi6yGr1Qq9Xo/a2lps3bp1wstACEkOiaqH7HY7zGZz4Nhutxv5+fngeR5FRUUTXh5CpqPp04WHk75kHHDbew6dPX0Rq6ZpNbhw+w514yEkTG1tLdasWTPhx7VarRE3Kzt37kRlZeWEl4UQkliJqodMJhPKy8tRWloKo9E44ccnhCSPRNVDLpdLErgxGAwoKSmh6yFCJtC0CaB879ndgddMpQkAB5/QjxMXrkdd3+v1gb/vmtAyEpLsnE4nNmzYMOHHrampiXjSYzKZ4HA4JrwshJDESlQ9RAghokTUQzzPo6ysDE6nUzJfbIkbPp8QMj6mTRee4heeDPx+vuk6Djp/Bq/fh7O3zmHjwysi1tdrNPjkyi0snpM2gaUkJPk4HA7Y7XbwPA+ANR+12+0oKiqakOarPM/D7XZHHMtgYF3snE4ndeMhZIpLdD1ECCGJrodMJhO2b98ecSy32x1YTggZf9MmgBLqxSeewFvnfg74gBs99ejzfhEqhfStMKSm4ErTXXh9q6GQT5uGOoREsFgssFgsqKiogMvlQnl5+YQeX7wwiMXlopZihEx1ia6HCCEkGeqh0tLSiHmHDx+G2WwOPFgihIyvaRlA+fzDS2GQz0RzfxO86MInl27j8dWLItbz+f3g77mwZF56AkpJSHJpaGhATk5O3Ovv2rULHo8n7vXLy8vp6QkhZFBUDxFCEi2Z6iG324233noLhw4dinv/hJDRmZYBFJmMw+cXfBFHPrwPo+cJeJQzgNUtEevpU1Nx8mIjBVAIAWu6OpwM76Ej5hBCyFigeogQkmjJVA8VFxejqqqKAr+ETKBp2zfllY0vwnjzJai7luLKOR2EKAPu6FNScON+M/r6fRNfQEKSiJiYzGKxTPixxSapsbry0GgYhEwPiayHCCEESK56qKKiAkVFRUlRFkKmk2nZAgUAHn5IhlSNAH8/0OFW4O4tNeYu7I1YT8ZxeP/sFWxauxwyGZeAkhKSeA6HQ5Ko1e12D9nXdteuXcM6RmlpadQnKCaTCQaDAS6XS3JMMaBCCWQJmR4SWQ8RQgiQPPWQzWbDhg0bKHhCSAIkRQBl165d2Llz54huhNxuNyorK4edyEkuB5at7sKFsypwHHC5QRs1gDInzYjPbtzEnTYXXrA8BINWPewyEjLZORwOSX9fm802ZPPVsWyympOTA6fTKbmgqK+vH9aFQ6zRekIz6vM8D7PZjIKCgrj2GbqtGOihixlCxkei66GxQPUQIZNbMtRDDocDmZmZkroktA4ghIyvhAVQeJ6H1WqFXq9HbW0ttm7dOqL9vPbaayMuw9q1PjR84gfUbpzkP8OTWBh1vblpRri7umE9chxfffwRLF2QMeJjEjIZ6fV6LFrEEi07HI4JvzgvLS1FcXExcnNzA/MOHjyIkpKSmNuE1jFA9NF67HY7zGazZL/5+flwu91DXhBZrVYYDIbAena7HWVlZTh69Oiwzo0QEp9E10MjQfUQIVNLoushh8MBt9sNk8kUCJoA7Joo2gg9hJCxl7AAislkCrQaOXLkyIj24XA4IiqQ4cjOluMHH/4NegwNAARcvfMXeHBe9OCIQZOKVJ8KBz88iceXL8UXHllKXXrItLFz507s3bsXRqMRBoNhwrvNmEwmlJSUwGq1Bv7mt27dOmg5QuuYffv2RV2nuroaOTk5kouOnJycIZ8o8TwPh8OB/fv3B+ZZLBYaWpWQcZToesjtdmPv3r3weDzgeR6VlZXIyckZtBk91UOETC2JrIfcbjcKCwujLjOZTBRAIWSCJEUXnpEQ8x8sWrRoxAEU02w9DEagZyCD7IlLF/DgvNjDkinlciyeOROnr1wHAGxau2xExyVksjGbzQlvCm+xWMb8SU9JSUlEElqPxzNkE9iysjLJ02IA1GyekHGW6HrIYDAEblDGMkhB9RAhk0ci6yGDwYALFy4k5NiEkKBJOwpPTU3NqC8StKkqrJgRDIKcb3XGtd2CjHScvd5Io/MQMslZLJaIZHA1NTWDdg0CWP6VzMxMAKwl3EiDuIQQQvUQIYQQMnlMyhYoDocDeXl5w9pm5cqVkulXXnkFhdt34tGli3G8WQYBfrT4r+FuawfmZOiG3J/PL6D++l2sXT5/WOUghCQfMQlje3s7Dh06NOiTX7fbHWgBZ7fbYbFYwPM8ysrKUFJSEjMb/1OdnShaKM2z9Pzzz+PLX/6ydMXUVCA7G9BqR3dShJBJZSLqoQ3x1kNKJbBuHZCePrqTIoQQQqaYSRdAES8ahhoyLFy0Jm9+vwCjRo3Z8uW462PLHRcv4YUnHh1yf7MMehw/dxmPLJ1HuVAImeTEZvB2ux1Wq3XQG5DQkTLE7cxmMywWC1577bWYTXu/2d6OL3z+89KZN28C//zPkSs/8gjw+usjOxlCyKQ0EfXQ7w+nHlqwAPjhDwHFpLtUJIQQQsbNpOvCU1NTE9Hnd6RkMg6pahVWz1oTmFd3tz6ubZVyObp6+8Dfj8yoTwiZnHJzc2EymbBt27Yh1w2/sbFYLDhy5EjgqXBULld8r9pa9pMQMu0kTT104QJw/vwoz4YQQgiZWibVYwWn0znmydHSNBo8tnQZ3m/iAAi46zuP63fa8MC8oZutpuu0OFZ3GX8wZ/2YlokQkjgWiwWVlZVwOp1Rs+uLNyzhzevF+fX19VHrqd9ptfhCQcHgBz99Grh1i9281NUBTz45wrMghExm41UPHYunHrp0Cbh4EWhrY8HcgTwrhBBCCJlkARSe51FXVyeZV1tbC7fbjYqKCmRlZQ27dUq6Tgt3VzcWKDNxq5/t+4jzNHbO2zTkthqVCjdbW9Hq7kaGIXVYxyWEJBbP88jPz4/INSDegPA8H/XGRVw31hPeWHkLfpaWBlnrn0rmpaUBL74IzJs3MOPAAeCttwCfDzh2jAIohExxE10PHUxLw1//6Z9GXRbw8cesC6HPB7z7LjDIUMqEEJIsdu3ahdraWsmoZmvWrEFpaWnMOtFut8Nms6G+nvVAyMzMRHl5OUwmE8rKylBTUyPZn8Viidm90mq1oqamBjzPIzMzEyUlJRM6zDWZOJMqgJKbmxsRILFarairqxvx2Odz0vQ4f/M2nn7wcfzsAgugnHN/jK6ep6FJUQ65vVatxkfnG5H32Moh1yWEJA+3243MzMyI4UPF3AKD/dPbvHkz6uvrJes4nU4YDIZBEz/W1kbOa2kB/s//GZgwmwGZjOUcOHMG6OykZLKETGGJqIeG9NBDLImsSgVcuwZcuQIsXTry/RFCyATJy8uTDDNvtVqjBqkBoKKiAm+99RaqqqpgsVgCI6CFeumllyT3mGVlZcjPz8fRo0cl6+3atQsejwdVVVUwmUxwOBzYtm0bXn/99TFLPUGSR1LnQHG73cjPz4fdbo+5Tnt7+6iOMcOggdfvR87qB6DFLABAPzpx7LOrcW2frtWivvEmevq8oyoHIWRiiQkXw1mtVhQUFEie8IbXQ6WlpbBarZLt9u7dO+Swo9HSDJw9CwjCwAqrV7MACscB9+4Bp06FLCSETDWJqIeGlJLCgigqFauPfvUrqocIIZNSUVERXnrpJRQXF0vmO51O7Nu3D4cOHQrUwQaDQVLvRiMGZxwOR2Ce3W5HQ0MD9u/fH9jWYrHgwIED2L179+A5qciklLAWKG63G3v37oXH4wHP86isrEROTg42bNgguZjgeR6uKMkUeZ7HwYMHceTIEbhcLlRUVGDLli3Dbiql16jBcSyh7PrZT+L0+bvQNz+L5qYlELJvgYtjgB2ZjMPZK3fw+OpRPPEhhIyp0DoGQNQ6pqioCDabDY2NjUhLSwPP87BYLCgKa7IeXg+ZTCbs378fFRUVSEtLQ3t7O7Zs2TLkU4Yf/CD4e2UlcP8+0N4OuN2A0QhApwMWL2a5BwDg3DlgxYqQPj6EkMkkGeuhuKxbx5rMdXcD9fUsN1PY8MeEEDIZlJaWYuXKlZKcUnv37h0yWBKLy+WStBqsrKyMGrg2m83IzMyEzWaLqM/J5JawAIrBYAg0iQptahW+zsmTJ6MuM5lMKC0tHXHXHZEuVQUxSvL82sdw5+gD8Ho53OsAmm6pMW9h75D7mKHX48SFq8heuZCGNCYkSYTWMTabDYcOHYq6XsEQCRVj1UNiHTQcjz8e/H3BAhY86elho4gG/hevWcNuWPr6WISloQGYOZM1qSeETCrJWA/FJTsbkMvZ742NLKn17NmsVQohZNr41a+An/2MXatMtJQU4JvfBF54YfT7MpvNki6PDQ0Nw26t53Q6sXv3buTl5Uke2MfKVQWwHCzh+TvJ5JfUXXgmQopKAYWMvQ2pGj9WZnUElp39yBhrMwmFTIbu/j5cud06LmUkhEw9CxYEYyK3boUsMJtZDhS/n+UfUCjYiBiEEDJR5s4FFi1iv3u9AM+zYY0JIdPKr3+dmOAJwI7761+Pzb6MRiMaGxsD0zzPR00EG27fvn3Izs5GdnY28vPzA0lmQ/cj7j+atLQ0NDQ0jLL0JNlM+wAKAKSqVPAP9O995PFg89gLdTp0d8X3FmXodDhzmR+X8hFCph6TKfiA9+bNkAVmc3DB9evsEcytW6yfDyGETJTHHmM5UASBBXPv3KF6iJBp5qtfZZchiZCSwo4/FlwuF9LS0gLTJpMprtwk27dvx8mTJ3Hy5ElcuHABJpMJ2dnZgcCJ2AVInA7H8/zoknqTpDSpRuEZLxlaLVzd3dCoVJi7sBdzF/Ti9h0BrWnHYT/dgxeeWjXkPjQqFZraI3O1EEJINKHpBEIeigAzZrCcJy4Xe/J78yZb+bPPAIuF3dAQQsh4W78e+Jd/Afr7WVfCr3yFdeWxWBBXgjhCyKT3wgtj04Um0ZxOp6TLzpo1a+BwOIadM6qoqAg8z6OiogJ79uwBgMCoO9G68dTX1yMnJ2d0hSdJh67EAcxK06O3vz8wvXD9J2jM+hM0L7LCcb8Gfn982ef7vF509fQPvSIhZNoLDaBcuxa28KGHgr9fucK68fT1sWb0hBAyEdasAQwG1p2wrY2NuS4mbSKEkEnCarVGjHi2c+dO2Gy2mC1HBmMymQKJwQEWVKmuro5Yz+FwwOl0YufOnSMrOElaFEABMNuoQ09IACU7Kw1yOXu60oVm1J67Htd+BAi47+ocjyISQqaY+fODD3Fv3WIPeQMyM4MtTa4ODKluMLBcKInqjEwImV4UCmDt2uB0QwPLdn3xIgvoEkJIkrPZbKiursbrr78umW82m7F9+3YUFhbC6XQG5tvtdslw8eEcDgeqq6slo+oUFBQgMzMT+fn5gYCMw+FAcXExysvL48q1QiYX6sIDwKBNgRDSykSTosQaw2M46/5vAMA7l45h3XITUlSDv11KuQKN99uxeE7aeBaXEDIFqFTArFlAZye7F7lzJ5izEWvWsDwoYiJZQWDRltRUdhMTelNDCCHjJScHOHKE/d7QAHzhCyz79YULQFZWYstGCCFhampqUF9fH8hvsmbNGhw6dChqHpLS0lJkZWVh9+7d4Hkebrcbmzdvloxs9tZbb6G2tjawP4PBgKqqKklrFgDYv38/rFYriouLwfM8MjMzo65HpgYKoAAwaNTgwvrzbs5ch7MOFkC577+EipqD+PPNL0GbEnsIP0NqKm7cbcFTmQ+MZ3EJIVOEycRan/T1sVbxgQDK/PlAejobxri7m0VX5s9nAZSWFjZ/1qyElp0QMg2EJpK9epXVR1ot0NTEKqwYI08QQshEE3OSDEdubm7MPCjl5eWSEXeGUlRUJGmZQqYu6sIDQJcaGRRZPDcd69KeC0w3+c7h7+z/gjZPd8z9qJUK3HNRhnpCSHwWLGCt5IGwtAIcx7rxDIwOFujGA7AbFqcT8PkmrJyEkGkqPR148EFWFwlCcChjgwGor2et5AghhJBphAIoABRyGVKUyoj52575HJ7I2BKYbvFfwxtH9+Nua0fMffkEP9ydveNSTkLI1LJwYTAPSkQes9Dm8aEBFLmc3chcuTLu5SOEEDz+eDBgK+YKUChYPqZbtxJXLkIIISQBKIAyIFWtgjfsSYpMxuFbz2zAxrn5gXlu4RYq37fiRlNb1P34/ZRIlhASn9CReG7cCFv48MPB369eDbZGAdjT3+vXWQIVQggZT6F9+M+fD9ZFRiNrkUIJZQkhhEwjFEAZMFOvQ3eMi4B8yzo8v+ib4Aberi40418OX0J7S2QKmRSVEjfuRQ+uEEJIqAULgr9fvy6NkWDlSkCtZr+7XEBrq3RjrRa4fHm8i0gIme6yslh9AwAdHUBjI/ud41g27IsXE1c2QgghZIJRAGXArDQ9+rzemMtz15tRsHwbZFDCeC8PqVe34hc/WQBXmzSIYkhNxbWm5vEuLiFkCkhPBzQa1ivH7QbaQmOvMlkw9wAg7cYDACkpLKGsJOpCCCFjTKkEVq0KTjc0BH/XaoG7dyknEyGEkGmDAigDZht16O2PHUABgKeylmJnVglmN30LHDi4XQr8+0/mw+OSB9ZRyuVo7+qE3083NYSQwXEc68ajVLL7j4h0AqtWBQMk0XKe+P3UjYcQMv7Wrg0mjA0NoIjaqOUtIYSQ6YECKAO0qSoAQwc9Mpfr8NXfa4JcztZtb1PiF/vno6U9mDjW7xfQ3tEzXkUlhEwhYgBFEMJG4gGAhx4K3rSEt0ABWNOVlpZxLyMhZJrLyQn+fvMm61Yo0miC3XoIIYSQKY4CKAMMGjU4cTiMITywvBtf/kYTZDJAgB8XtD9GxXtWtAwkjxUA3G2LPVIPIYSIFixgcRAgSgDl4YdZVx4AuH+f5R8IpdEATU3jXkZCyDQ3bx5gMgWnz58P/q5SsRYog3SDJoQQQqYKCqAM0KiV8TRACXhwVRe+WNCE5gf+P3TMOAYP7uAH772Jnj4vNCoVGu9Tc1ZCyNBCR+KJGMpYr5dmmg3vxqNQsKAK5R8ghIwnjYYlthZF68YTnuiaEEIImYKGDKCcO3duIsqRcDIZh1S1aljbrDB34gtrFwNgLVfaBR41Z+qgS03BjXvUrJ4QMrTQh7oRvXRSUoAHHghOX7sWuQNBADye8SgaIYQwKhWwZk1w+sIFaYsTrZa68RBCkobdbkdhYSGys7ORnZ2NwsJC8ANPqcrKypCdnY1NmzYFXmVlZXC73VH3ZbVakZ+fH9iP0+mcyFMhSWjIAIpOp8OPf/zjwHRtbe24FiiR0jSaQUfiiea5tauRM2NLYPr47f+G3++Hu7sbXp9/rItICJli5s5lyWQ5jvXGkYymrlCwkXhE0RLJqlTAvXvjXk5CyDS3ZAmQlsZ+7+2VBnSVSqC9HejvT0TJCCEkoKKiArt370ZRURFOnjyJd999F7m5uZJ1XnrpJRw9ejTwAoD8/PyIfe3atQsOhwNVVVU4efIkioqKsG3bNtjt9gk5F5KchgygmEwm2Gw2vP3223j55ZdhCn1cGkVHeB/9SWRuuhGenuEnf33h8fVQCjoAQDda8PaZc/ALfrS6u8e6iISQKUapZEEUlYrdk9y+HbbC8uXB32/eBMLrKI2GDSNKCCHjKSMDWLYsOB3+FFYmo6TWhJCEcjqd2LdvHw4dOgSLxQIAMBgMKCgoGPQetry8HADgcDgC8+x2OxoaGrB///7AthaLBQcOHMDu3btjtlghU58i1oLa2locP34cDocDjY2N2L17N959913odLqIdd9++22YTCasXr0aNTU1+PrXvy7Zj8lkwsLQjv5J6rGVJpy5en3Y22lTVFg/82nUtvwWAPDBzffwyNLfR1ObB7PTtWNcSkLIVLNwIWv93tvLYiShvXYwezYwZ04wSHL9OhveWMRxrNlKby+gVk9gqQkh00p6OgugnDrFps+dA7761eByrRa4cYNFhAkhU8bP6n6Gn9f/PK51Ny/djFcee0Uy74cf/xBHrhyJa/tvZH4D38z65rDLKNq7d++QwZJYXC4XjEZjYLqyshIlJSUR65nNZmRmZsJms6GoqGjEZSWTV8wWKDk5OSgpKcGhQ4dgMpnw9a9/Ha+99lrUdQVBwI9+9CM8++yzsFqt+P73vx/o6pOTkxPoc5bsDFo1HnlgEVo8w29F8+X1a6GABgDQiXs4fekWrt+lhGqEkKEtXMh66wBRRuIxGIZIlDKAnoQQQsaTRsO6FIqV1b17QHNzcLmY1FrSD5EQQiZOQ0NDoOVJvJxOJ/Lz85GXlwez2RyYz/O8ZDrUmjVrUFdXN6qykslryC48Ho8HmzdvRklJCV5//XW8/fbbEets3rwZVVVVeOedd/Dcc88hMzMTBw8exLPPPouvfe1rk6qf2JOZS9Dn9cIvDGNIHgAGbQoeTf9cYLr29gdovE9NWQkhQ1u4kDUkAaKMxKPTAYsWBaejBVBSU4E7d8atfIQQgpQU1sptxYrgvGij8YQGVQghZALxPA+DwTDkevv27QskmM3Pz0dmZmagG4+4HwCSFimh0tLS0BCt/iPTQswuPKL6+nrs3LkTAKDX6/Hcc88Nuv6WLVuwZs0abN68GQALwOj1+jEo6sTQpCjx+IoH8fHla5gb448mlq+sX48z7/w3fOiFB3fw3mcXsSPPApVSPk6lJYRMBaE9HK9fD1uYkiINoNy4wUa/UIRU32o1yz0gCMFIDCGEjDVxOGPxxuHcOeBzwYdH0OlYf8T58xNTPkLImPtm1jdH1a3mlcdeiejWM15MJlNcuUm2b9+O0tLSwLTVakV2dnag54XYBShWKxSe50fUTYhMDUO2QMnJyZEEQH7xi1/gxIkTMddfEzrMHTCpgieix1aZwAHDboWSrk/Fw8anAtOnr1/AfVfXGJeOEDLVhAZQbtxgcZAAtRowGoGZM9m01xulmQoAnw/oovqGEDKOwhPJXr7M8i+JFAqgs1M6jxBCJsiaNWskiWDjVVRUhLy8PFRUVATmmUymmPuqr6+PuOcl08eQAZRwdXV12LZtG1avXo2vfe1r+P73v4+33357Uo++Ey5FpcCTq5fjnmv4OQW+uv4xpHY/iFnXX4Hm07/Ae8eGP6oPIWR6MRgAvZ7de3g8YQNZyOVswZIlwXnRhjOWyYBWyrtECBlH6emsFYqYKNbrBS5dkq7DcTS0OiEkIXbu3AmbzTai/JsmkwkejycwXVRUhOrq6oj1HA4HnE5noIcGmX6GHUApLy/H+fPn8e///u/Iy8uD0+nErl27kJ2djeeeew7/+3//b5w/f348yjqh1q1YALlMBq/fP6ztZhi1+FL6/4C+9Uko5Urs/6kf//P7ddj6xr7A609+aIPt/bPo6KZEa4QQZuFCNpSxzxclkWxKinRonmvXIneg1QJNTeNZRELIdKdhyfIR2qQ9PA+AThelEiOEkPFnNpuxfft2FBYWwhky1Lrdbh80J6fD4UB1dbVkVJ2CggJkZmYiPz8/EJBxOBwoLi5GeXl5XLlWyNQ0ZA6UWMxmc+BLCrAMxtXV1airq4PNZoPZbMZPf/rTqMMeTwYKuQyfz1qJdz5twLz0tGFtu/7Jdpz92Ah4Zbh3RwGbvREty34hWedNJyCzKZEuM2FZxjI8sWwVnslcgSfWLIYuVTWGZ0IImQwWLgTq69nvN28CjzwSstBojEwkG57vRKEA2ttZBEZOeZcIIeMgNZX9XLUKePdd9vu5c9L6SC5no4L19LDgLyGETKDS0lJkZWVh9+7d4HkebrcbmzdvluQ8eeutt1BbWxvIl2IwGFBVVRUxgs/+/fthtVpRXFwMnueRmZkZdT0yvYw4gBLObDajqqoK+/btw6FDh2C1WvHtb38bv/zlL8fqEBPuoQfn4ncNl9Dv80E5jBsSncGHhx9z4bQjDUqFHEqFHDJZtMSOXrThGj5uuYaPmt9G1QlA27MCn+//PlYsVWLuXA5yOXCx9wM0+VirHjknw2x9OhZkzMDiWTOxZO4MPDg3Aykq9lHKOBm4gYsYuociZPJYsCD4N3vrVthCg4E91dXp2DChPT1s1J3wRI2CwJYPMwE2IYTEheNYK5TFi1lwpKeHBW5v3ZImc1IogLt32XqEEDLBcnNzkZubG3VZeXm5ZMSdoRQVFUlaphAy7ABKQ0MDbDYbMjMzkZOTg4Wh/zARHO6pqKgIFosF3//+9/EXf/EXY1PaCSaTcchda8YvHKexMCMDsmGMbvHU5hYYM/rRdN+LTk6BWYrPAwAEQYCrrw0tXh49HMtXwHEIBD24znn49KoLdRdkMGpToVIocHP+KbSlvRvcOQdw4CDjEDHihql5G+a1fw0Au8aZMQO4N+MQ+rXXMVs/A6lKNeQy1nNLxskg47jAsQFgueYxzFBKP9NPPHb0+rugVCiwID0DC2dm4ME5M/DA3HRqLUPIGAmtSj/+OCwA2pUG9e25eFgwY3bLachlwL2jt+B6eLl0Jz1yPKhpwVyLEYIAuFwsKa04qqhMxqoM2TA7b65eDcyaNaLTIoQkmcJC6bRCAWRmAk88AaxZA/j9LAes388anKhULJd1oN5IT2eVyurVwCefsHlHjgAvvxzcqVbLkl1TAIUQQsgUM+wAysGDB2EwGPDzn/8cZWVlWLNmDSwWC7KysgAAx48fx9e//nUArFVKbW3tkPvctWsXdu7cGXWYqGjsdjtcLhcaGxvR0NCAgoKCmFHG0Vq+cAaefWgNjn7WANOMGXFv1+PtxawV1/HFLQ9Cn7oKwKrAsj6fDzfuteJcYxPO32xCY1sT2nrvw+W9B0X3PACAz+/H3XYXC7gYOtGtYflSBADBcAfHgi8IBkHuuTzovcsCMzKOg6yRQ9PyWnQb6sHiLYMHgRY0GmBwZUjmXV55EP2qe4F9chzHgkkcoOJSwYWk0nno/t9A179Usn3t/G9AgAClTA2VLAVqOXulKlIh46R3cjn4M6i54MhNvYIHtfj7Qcsc6mnslpyjR7iD07BGXTf8vVBBBwv3Z5J5LcIl1OFgXMfWYx7Wcdsl824LZ3ARv41r+xlYjixuq2TedeEDXMexuLafh0exkvuSZN4F4be4gzOSeRyifwcWc09hCfe0ZF6d34YWXIq6frjlXB4WcOsk807596ED8eXlyORewkxuhWSew/8P6EN8CarXcS8jXTkPn/888PjjkcudTmfUOsZutwf6torD1RUUFAx5vLGuh8JH4pG0QvGrwbXNxhXPWuTdq2Pn8x938cvTcyP2o/y5F3/0V+zhcH8/C6SmpY24WPD7gU8/ZTdXk3BQNUKSSjLUQxcvRisXcPAgC5SsWsVSnKxYwQK54r9KhWJgULDeDGTcuoXU9V+A5uQnLLDyWT28F64CSx4EAMhkMih6etjIYGLeFEIIIWQKGHYAJTMzEy+99BIA9k/+4MGDOH78OGw2G0wmE773ve8BAH784x9j8+bNWBTabz8Ez/OwWq3Q6/Wora3F1q1bo64Xzm63w2w2B8bedrvdgeQ+49W8KnvVQri6enDm6nXMT08fcv27Lhe06hRsf24DZhqjXzhsMLOnMu7OXtx3deLGvTYoZDKoZCp0d8jRfFeBDo8WHAc09m5Cq3c1AKDP24+77jbc62hGc1czXH0t6BZc8Ass2a1cJoNCzkEQAJ9fQF+fF92yZnh9vrjO1dXVjX6P9Ia1t78fXpk3Yl0OHHpk/ZJ51+64kNLlku5zRjsECIHAC8eJgR+2F4krPVB6g+9Zv6IHdUs/jqvsAKC44JMECLpS3Di/OL7tlb50+C9L3yeXtgVXFn4U1/aa3qXovi7dvjntDhrnxLe9sdMH103p9ndm8LgzM77tb7uMuN8k3b5xzlU0p8W3/b2WxbjZLN3+ysJzcGlPxbV9291HMbNduv35xXXoSokyYkwUHbc2wtgh3b5u6Rn0K+IbWabnxovQ9szGxx/L8K//yqGlJVjHAIDL5YrYRqxPQm848vPz4Xa7B61PxqMemjePdePp6GAtR6RkULmBOt9qbBr4k5vbfh6tqUJEKzS5T0DV3/vwP78rR0ZG+H6GTyZjPYhOnQJyciilASHDEXqtAyRHPeTzRVQbAAClkgVMGxrYKy0N2LwZyM4Otojz+4H2fi36mgT0pi7EnLnrYLh0GgDQ85PfgM8vDuw8c7ECmrt3pSOIEUIIIZPcsAMoOTk5geCIyWSSJOQJdfz4cRw+fDjmEE8mkynQ/+zIkSNxH9/lcgUuFgCW9KekpATFxcXj2j/tC48sRUdPL67cuYs5adHzC/R5vWhyufDY0iV45pGlUMiHbidv0Kph0KqxdP5gdzo5wy6vILDhUFtbgffPv4ILd27jRvM9uLu7AQjwCQIgCPANBF4gCACAhTNnwwjpTexF7mn0owN9vn64elxw97eh0+dCD9rhF/wQIATW7fX2s8feIXx+AeAEwC8E9swNdD8Kv4a77+6Aol8ZmPYqO9DnjQzexHLP5ZEEUHr7O+Pe3tffj3suj2Rel9AV//F7+yK2dyu7496+syfK9pr4t/d090Rs7zH2xL29q6sb8rDtO2f2oU8d3/btnd3wh23f3dePPkV827d2dKE3bPue/n74EOf2nk50dnmwelYKUlNTJHXMvn37om5TXV2NnJwcST2Wk5MDm802aH0yHvWQTAb87d8CDgd7aBvK7+fQ/1kzugU9Zh9SgOvtwUx/M55/6Do6UmcH1mu4lorW+0CX24uf/UyOHTui3ygNl0LBmvGfOQM89hibJoQMLRnrof/4D+n0/fvAb34D/O53LIAr1hnt7YDNBrz3HvDFLwIPPcTqKZUxFampHFR6oHfjFigaPwV8PuhabmB201l0r3gEXh/gvKHFQyoeagqgEEIImUKGfRlsMpnw8ssvo7a2NiL/Saif/OQnoypYNDzPo6ysDJmZmZImsOLvsZrGjgWZjMOXHl+Ntz7oxX2XBzP0bHShfp8P7u5udPf1Qa1Q4htPZWPJvKFbqYw3jmNPjQ0GYNsD6wCsG3Kb2HZEnev3C7jX3imZl6pIhVwmzV7b0ffv8AsCPF298HT3wtXZDXdXDzzdPfCFDRO94lsLoJQF86r0+zW46I4/0dOaP0iTdM3p9q7E1Y7I7f2CEDFPwSmx6uU0yTx3/6No7PzruI6dKtdiqV66fWvv53C72xR9gzB6RRoW66Tb3+vJw72erLi2z1DNxnyNdPvbXS+ite9zcW0/O2UBZqdIt7/R8W14vF+Ja/v5qUuQoZZuf8XzJ+j2dUbfIMwi7QoYlNLtz7tK4RX6o28Q5sFvroQhRYesrPiDBiUlJYG8TSKPxyO5KQk3nvVQRgbwpS/FWLi8C/C6gZb5wMBQ8X/5zAkW0Rhw0qnBd/9hDnq7e3H+vBrvvgts2jSiokRISWE3V599Bjz66NgEZgghE18PzZkTOZ2ZyXK+HjkCXL7MWpx1DDRGvX8fOHAAMJlYIGXlShn8KjV7CGPMgOfRp6A/9T4AwPi736J7aSYUcgWUKhku1vdj1fpOKNO0w3tTCCGEkCQ14ueIOTnDbxUxWiaTCdu3b4+4qBCHoBrsYmMsKOQyvPjkw3jzvVO41dYGOTikabXIWrwQS+fNwLwMPVTK6TP0jUzGYW7G0MNU6zCwzoyRJFBQYQMeG3q1mLR4elTbpwOIklAjbrMHXiO1YOA1UksGXiPzOFaO4tjA48gc5fZrR7X9UMKHoXO73aipqcGBAwdibpOweshgAJqagAcfDARQcPWqJICSbe7Ci8+58W+/YX9zv/0t8MADwLJlY1MEnY49lb54EVg5uq8GIWRAstRDc+YAL77IWpo9+ST7+d57bKAdgOWE/dGPgOXLga9lzcaDhvsQVClwP/4stPUfQdbTDUV7M3Sf1aLj0aegVgO9HQo0vHcHWV9dNuzk1YQQQkgymnQNsaN1GTp8+DDMZjMMBkPM7VaGXe2/8sorePXVV4d9fJVSjt/7/Dq0eroxO00bVzcdQkhyExM4tre349ChQ0PefCSkHjIYgMZGFkARXb0asdrLX23BiRPA1f50QC7Hm28CpaUs+DEWjEZ2I6VQsKJQSxRCxkYy1EMaDUsY7XQC69cDFgvw7ruse4/YG/TSJeDv6hdh/VwZvviFLsyboYHn8WdhPPYbAICh9gg6V6+HkJIKdYYWPfwt1NcvG1brQEIIISRZTboASji324233noLhw4dirnOH7e04Dvf+Y50pssFDCS8Ha4UjsN8jYYNSTHYKyWFrhYImQTE5I12ux1WqxUlJSWD3oCEi6ceAoALFy6MvJCpqSxX0aJFLKOjz8fa1rtcLKoxQKkEir/G43/9Yja6oYHbDezbx0YclcnYSxxZI57qSaUCHn6YHV6Ung5cv84OnZXFjkkIGZ1kqYcUCpbv5OJFFrP98peBz32Ode/56CNWDQlyBc5cNeITfiYWz+2BUngJ32o+A31fMwAfLu4/hQcLn4Y2xQ+tsh+3GztwKUWHFSsGPTQhhBCS9CZ9AKW4uBhVVVWDPql5qKeHZWYcS6F3HmI+jfCfcjkwaxZrP79iBXs9+CAbB1C8i6E2rYQkjdzcXPA8j23btg15ExIqnnpo1NRqVu8olWxUi8uX2fwTJ9hQGSGWr1Hhm5br+LFjDQA2LPKNGyM/9PvvA6+8Im3FkpbGciQ4HOxJtZZSHBAyJpKhHuI41k1PpWItTjIygIIC4POfBw4fBs5+yrorCwJw/U4KgBT8MuX38JL7HwEAc53v4cDPn0PhN3qhU6owo+82btxYAY1GOmQ7IYQQMtlM6gBKRUUFioqKIvoPR9XRMfQ6w+H3B1/iEMHi491Qra3AhQssGQHAHu3Mm8euIMTXzJnDG9ZCq2X7yMigMUUJGWMWiwWVlZVxJ4MdVj00GqpgcmU88UQwgPLhh8AXviBpBqI1KvDQrDt4dtMqvHN09EHau3eBvXuBP/kTaUsUrZYNuuVwsCfW4ckpCSEjkyz10JIlrGpxOtklx+zZwLZtAM9zeHtfF87dCrZ+O6u14EnVf2F+3zUohD6svPCf+OGhb+PVfAHa+7eRbloOp5MLDI2uVLIXPUcihIyFsrIy1NTUSJJyWyyWmK35rFYrampqwPM8MjMzUVJSMm6DkZCpZdIGUGw2GzZs2BDXxcKPMjLwhTfeGLuD+/1AZ2dwnOC2NqC5mf3e2sqWdXayDsMcF2wvL2pqYq9Tp9i0RsOuSuRxJKCVy1nwZPZs1pQ/LQ2YO5fduej18e2DEAKe55Gfnx+Ra0D8J8vz/JD/SIdTD40ax7FWKADrU/Nf/8UyunZ0AKdPs6BKyKoZ6QI2mjswd54B9+8HY70+H3tqHDYAVlR9fazJPgDcvAlUVwN/9EfBYgDsBigtDTh7llVBYtcg8TV7NrB48Zi9C4RMKZOhHlq4kP2df/op674nl7MRef70m21oaboPd38wqqq/8yzm/eaHuNOiwlrP+zh+PQ//9Kv5KH62HfIuDzIyDGhokF4SyWTsGZJazZ4JabVs/6ExY0IIicdLL70kyQ9VVlaG/Px8HD16VLLerl274PF4Aq32HA4Htm3bhtdffz3QnZKQWCZlAMXhcEQM3cfzPIDomefPpqZKbi7Gnd8P9PaywMrZs+zm5sYN9hi3vT1y/a4ullAgXhcvBn+fMYNd3cyfz+5Sli0DFixgj4pCHxUTQiTcbjcyMzMjhg8V65KhblqGWw+NCZ2ODYmhVAJPPQX853+y+e+/Dzz+uOSuJGOuEvy9+1i7Nv4cCtE88ABgs7Hfr18HfvITYPt2ad4TmYxVRYIQ7MEIsN8vXWIxX7oZIiTSZKmH5swBsrPZ5YzBwAIeXuMMzPFcxixtSDRkwWLoGpdh/tnLuN2sQm7rz/BT9f/CPx5Zjj+a0wTZCgPS06Mfw+dj8eCWFpZfae34DsJGCJkGysvLsWnTJjgcjkCQ2W63o6GhQRJUsVgsOHDgALZt2waLxTKs/FNk+knqAIrb7ca2bduwY8eOQDTQ4XDA7XbDZDIFLhIA4ODBg1Ez0ieETMaCF6mpLLCxeTMLnFy/zoavuHmTXSHwPAusiGMEjkRLC3udPcum1Wp2zEWLWCDFbAYyMxHzioWQacpsNkd9Ymu1WlFQUBC4+UiqekivZ3cWSiWQk8OyOvb1seDshQvAqlWBVXUzNUi90gRg6agO+cQT7BC/+hWbvngR+OlPWTP+8J6H0RLTKhSsygsdPIgQwkymeigjg42afvIki+XKdAbA74tYr/1zX8bc62+wYGrLp1jaXYfLTVmo/qkXLxe3QD4rI2oGa7mcvdRqdllz5w4LvhJCEuhXvwJ+9rPR3auMVEoK8M1vAi+8MKrduFwuSZC6srISJSUlEeuZzWZkZmbCZrOhqKhoVMckU1vCAihutxt79+6Fx+MBz/OorKxETk5ORDNUnufhcrkC2xQWFkbdn8lkSp4ASjiZjF15ZGSwISvu3mWBk+5udiPU1QW43fHtS2ytcu0aC8D4wi5eenvZsmvX2ONfsb1+RkawhcqSJSzIMm8e6/5Dj4bJFBRaxwCIWscUFRXBZrOhsbERaWlp4HkeFosl4h9n0tRDBkNwLFGxZd0HH7Dp//5vSQCFk8swU9ONm55+pOhHN0zO5z7HqpbDh9m00wn86EcsNrxs2eCj+eh0rMpavJh6GJLpZ6rVQ0Yji91+9BEgKDTQIPKP3ztzHjrNj2NmPev/l9fyb/jhgv+Li01p2L+nHX/8vBNYtBjemfMgKKNff6SlAQ0N7NIltMsgIWSC/frXiQmeAOy4v/71iAMoTqcTu3fvRl5eXkQrvVit+9asWYO6uroRHY9MH5wghDa4nppWrlw5uuFDx0tHB3DrFnvFk5BAJD6ikcnYo93r14GrV9nPWMlyQ4MpHMeiumo1ezw8GRLRcpx05KJYP2WyxA4dLT5CE19i2cLnkQm1ct++hNcBY1IPdXSwUXfEFmUtLdLh2EtLWXBUXJ1vwynvI9AvmTm644JVHb/9LfDuu9L5ixcDmzaxxm6x/vTa24E1a+hpMpnekuFaZKzK0N0NfPwxkF53DMo0XcRyuacdc3/yN+C8XtxvV2CP/M9xVv8kAGDZwm58eW0jls/2wJcxC33zl8Cvi2wu39XFArCPPjrq4hJCBgy7DphELVDKyspgs9kC3W/cbjcKCgpQXl4eWIfneWzatAknT56MmVjWZrNF5EwhJFRSd+EZKxcvXgQXcmUfK2ZUXV2NnTt3BqaLiopQXV0ddd1169bhzJkzgelTp05h3bp1EeudPn0a69evD0yvXbsWp0+fZhM6HRsncPlywO3Gjl27YH3zzcC6e7//fez49relOxQEoLsb3KJF0tnvvMOCIZ2dQGMje92/j+qTJ7Hzs8+C5zR3LqqXLWMVYVcXu+NRqQCOw7rTp3EmJABzau1arNPrI8/J48H6kHNfq9PhdJRzB4AdFy/CeudO8JyWL8eOkBu8UNyxY9JzevrpKCtxqL59GztD8sAUzZ+P6jVrAstDrTtxAmcGnvwBwKnHH8e6KBXmabcb68VsmQDW6vU4HSNvzo6GBlhv3Qqe0+rV2CGOyygef+And+SI9Jy2bIm6z+rGRuysrw+ek8mE6qysqOuu+/BDnAlpsXRqwwasC+s/DwCnXS6sP348eE4GA04/+WT0c6qrgzWkCfjezEzsCPuOiTixGcKAyXBOK1asiLrOpCMOZSyaMYMllBW78B07BnzjG4HF2pmpSLlwF8DoAygcB3zxi+z30CDKjRvAj3/MGrMtXcqqodDX6tVs2ZUr7Gci45uEkLEhNoBruKJHn7sfcrX0clKmToNn3TMwfHQUs9K82N77L/gz/2PwylS4fDMVP7i5EvNn9eHpNffxxAOnkKpXo3fhUnjTZwWaqmk0LD//3bs0whchCfPCC6PuQjORtm/fLmmBZ7VakZ2dHUjULXaLjNUKhef58ctjR6aMaRFASXoyGWuvGt4SRKdjN0jxePRR9pi3vZ1tt3w5u1PR64GQAAoWLmSdmNvagiMIid2AwgNLPl+wu0D4/FCCEH09ILJljd8fe91w4euJGSrD53u9sVvehB+/szP6HVxnZ+R2IYEXif5+6XRPT/R1ox3H5Yo+Pzyy39cXu1tX+Psf7zn5fLH32dcXWZ54u5XFWi/ZzmkqiDbm5zPPBAMop0+zKMdAkJBLTcEceTNudY9NTmmOA770JXbj9N577Am0+NGJg4uF++1vga99jbVQaW+ndEyETBVqNZD5VDruftwIb4q0FUpnJ9C0ZiN0n9VC1t2JxeomvDrvl/j7O8EA7+37Kvz82AIcqp2Hx1e78MzyK5g/owF9sxfClzYTgkKJ9FQl6j9TIf1pGfU2JoQMW1FREXieR0VFBfbs2QMAgVF3ogVQ6uvrkZOTM9HFJJMMBVCmihkzpMEWn4/d6IeO2AOwQMv3vx+c9nqDAQGLhY1TKPqnf4qeBv/MGSD0qf/y5cA770Qv15/+KbB/f3D6O98BXn45+roajXQ6dJ9+f/D1k58Af/7nwWWbNgH/9/8Gl4f68pdZwgbR977H8tCEq6sDvvrV4PQDDwD/9m/Ry/mXfxkclgRg5/ONbwSPL44V6/ez1gGh/t//iwwWAGw0ldDPJTsbiJLgCgCwY4f0c/3jP2YtmcJduACEtKjCggXS7h6hKivZsLiir3wFeP756Os+84x0OtY+k+2cpoqUlGBXPIB9VxcvZk1BfD7gww+BkFZBM9O8uNrejdQxHJVr5kzgpZdYDpT33wccjsHjVb/8JYsx6vVssCBCyNSgnJmGhXOuAGGBUUEArl9PwZ2HNmPBR4cAAE+3/gr6r2XhvUuL8PE5Pfr6WR3W2yfDB2fT8cHZdCxb2I1nVt/Fo4vOQC4DAA6qXuDGNRmWrx4Y6zg1lV0vaDTS4cAGLaiSJXAhhEw74Ym2i4qKUFlZGZFnyuFwwOl04sCBAxNcQjLZUABlqgrNuTEYsZ09EPlkOyUl+mPr8JYy4qhDsfYfSqWK/1F4rPW02sjpGN1NIrLPLVwYfTiQtrbI7WINGxLeBWjWrPiHGInR3QTh/VHnzQOidWECWAujUOvWsddQ6+l0sfcZHixasSL2uuFirZds5zRV6HQsGhH63X7mGTY0DgAcP86CigOPa7U6wNDZhvb2kQdQZLLIrz3A7ke+8hV2uHPnWK9AseGa18vmidcsb7/NGr2ZzZEfIyFkktJogg8PQq4hOI7lq7/5XA666j6ApqsZst4erPuPMqxek42vv/h5fNi0DB+cNeJuazAIcvlmKi7ffACL5szDt/PuYt6MfigBtLqB1h4gQ+1j9V9bG6tk4k3j198PPPII60dICJk2HA4HqqurUVVVFZhXUFAAu92O/Px8VFVVBVqkFBcXo7y8nIYwJkOiJLKEkCkvGeqAMSvDjRssYXRoFMLnA/7mb4KBwBdfBDZsYL/397OnrzHyFA3F7wfq64H791lPw+Ho7WUNxsTGRT4fiyH+7d/SiDxk+plS9VCopiZWSaSmRk1K3/a7egg//jGUYc9Teh5cA/faZ1CHLHzwmRGfXtZBCGlEqlQI+OpTLXjmURcEgQVoH3kk/kYnEoLAkm6vW8ea0BEyTSVDPTReysrKUFNTA5PJBPdA926DwYCSkpKYw8XX1NSA53lkZmaiqKgo6nqEhKMACiFkykuGOmDMytDayrrRhScTOXaMDfcHsFZR3/1usJtPSwvrvjdr1ogOKQjA+fNswLDh5jDxeoE33wymYurvZ2mYxAfB4sBZoa/xJpezhLdmM+sBRQNjkYkwpeqhcN3dLBdTV1fUrjLu2np0/tIObdutiL/x/lnz4Fn3DO4sXI9j9TPxzsk0+HzBlVYu6sYf5N5FisIHnS74Nyu+4ub3syDzY48NPxpMyBSRDPUQIZMdBVAIIVNeMtQBY1aGzk6gtjYyktHTA/z1XweT9778MpCZyX4Xbxyys0eVxfXSJdb4JSNjeNv5fMAvfgF89BELqKSkBB9UJ3JUHr+fPTRfvpz1wlu0iMWYxC5IQ/13VCpZ1yajkfUkFEeHH69zEkej93qDqZZEoQOAia+UFAoOJZMpVQ9F4/ez4bauXWMBirBmZh0eAdfevYJZzvehve4EwvPWa/XoePQpXJr/NH7830tx634wa2yq2o+Cjfex2tQRsR0nA+L+k/N5oejywGN+Ar7UYCs+QWC9TFesoL8ZMrUlQz1EyGRHOVAIIWQySUmJfoeeksKGx3n/fTZ97FgwgCKO9HXqFHv6OsJkisuXs9QqFy6wIEq8gQK5HCgoYMU4dow9rO7tla4z0YEUMXe2TAbcuQP87nesDGo1y90w1JDLHMfSP+h0LHii07G3VaFg75FOx4IrMllwADHxJRKEyLzT4dNi0EScDhce5OE4aY5hlYol701LY+URe1nQcNJkzMlkrJKYORP45BMWYQxJDq/Tc1iWuwznFi2DovUeZp37ANqGj8EN/DHKOz0wfngY62V2PKLS4Ea3HtdbDOjj1OiTqdH/EzWaZgI6gxx+lRpQKCGTc+BkwpABlE7dHNxduB4+hQHwGcE1XkLPA6sBFYvkqlSs8Ux7O+smFKUnEiGEEAKAAiiEEDK5yOWRyZlFTz/NIhSCAFy+DNy8yRIni9sZDCyI8sQTkcmY47R4MbvZuHw5OPKOILDdi6Msi91y5PLgKOMcB/zhH7KGMc3NbDsxICAGCqIFCMZDRwfrbeBwsIY5cnnwqbPXy1raXLo0sn3rdKyRj17PghUKBXtfxFfotELBPga9nn00Wq30vRtRrocoenpY6pz+/mDrFLUaIxoWluNYWTMygq1uKBhDJNLTWbKjujr2B5aWFviSpKYADz0EXLs+G436F2HIyYPeWQvdJ7+DvGNgSHq/H4qeDixN7cDcmXdx854afb0DX7LGkRVJBSBF9l/4yLAJJwyb4ZHPBt7vgM+gBGSspYxOx0YX6+xkQZTQgQ0JIYQQEXXhIYRMeclQB4xpGU6ckN71h/qXf2FPfwGWMPH3f1+6vL+fNQF54on4R8SKQRBYS5KeHjYwRmdnsPuL+FIqWdAlIyP5msYLAhvl/MgR1vPg1q3IAbkmkkzGbuL0evbSatm957x5wPz5rHtRrNjZcI00YCUILHAmBs/EljgZGax8o+ghNuVNuXpoKIIANDayJmsGQ0REsLmZ/d2lpAAquReaC59Cd+YYVHdvStbzC8CdZhVaPWPz5fdxCnymteC4IRd3VIvh06dLKqeNG1lDvWXL2IsChGQqSYZ6iJDJjlqgEELIZGMwsDv9aO3Mn3kmGED55BPgS1+SJkxUKtmd88cfsyBK+FDfwyDm2UhJmZw5GTmO9XJas4Y11rl0ibVOuXuXBYQGIwiAy8Wa/Isvl2t05fH7AbebvaKRy1nXovnz2WvGjOi5TziOxcZmzozd0Ehs5TISCoWkZwZ8PvaeXbsGrFrFAmaEgOPYl2HGDOD0aRZtDRk9bOZMNnnhAtDRowDWrEfXmvXg+nrB9fdC1t8Hrr8PnLcPqr4++O9wuH1bgNDXD5mXLfP5AL8weIRD7uvFg3dPQNc90PQNXmzEe9joeQ9NxhVwap/Db7nn0dHF/iDefZflevryl9nf9cMPj6y1FiGEkKmJAiiEEDJBOvo6IuZx4CCXySHn5JBxMshlcdzVGo3sjjVaAGXRIjbEzJUr7I78d78Dnn9euo5azSIAYk6UseorMknJZOxtmzuXvW0aDXsN1dojPAjh9bLAS3s7i291dbEGP14va7Hh9Qan+/vZvM5Otk1HRzD/byw+H2slc+tW/OcmBlJmzZL+nDFjZK1ZxK5HoeTy4Ht24QKbFnuOEQKdjg2rfukS+/IqlYFASkoKkJXFGqrcaQL0OkCuUkNQqRHeQGrGYmDEvWr8G5ByuQ760+9Ddft6YPYC1GN901m8aHgTP0v5Q/xX9yb0y9S4cgWormYjwns8wNq1kzNITAghZOxRAIUQQibICf5ExHARXNgMAQLm6+dj5YyVsYMpGo10CJZwzzzDIgEAS/Tx3HORLU1SUtgd/unTwPr1Y9c3ZBJTqYDVq1kw5fJlFugYjNfLgh7iyDhi8tb09ODNltgiRExdI5fH7n3V388CKR5P8NXSAty+zV7t7cM/p+5ugOfZayzI5ayVydq1bBjo8K9VRgbQ0MDWmzdvbI5JpgCFgv1xPfgga95x8yb7I9DrIZNxeOABwJgGXLoIKFVAiprFf8WugD6fNDny8Mngmfcw7n3pYaTcvYG0z45Bf+VT1j8Icqhb72Cb7O+wxVeNpnshdeFJYOYsoCcD6NAlR6xZUKvhNz0A35Jl7PXAUvjnzg++OYIQfOMG+ulptdJWY5OSTEZJlwghSYGumAkhZIIY1Ia4Wpjc67yHlq4WPDrvUehUusgVhrqINJtZM4PmZnaH/9FHwOc+F7meRsPu2D/9lN0RJ1uSkgTRalmz/eESR9ARb/jEe5je3uiv8MCL2BUqVh6Rrq5gMOX2bfbRAcGReMScJoLAugE1NwdHGxorPh/LG+N0spvJzEz21Vm1it0jcxwLonz2GQuizJ49tscnk5xaDaxcyQIpPM+CKQCg1yM9TYZHHmENVbq6WEBTq2WtqDRaQDWQfJmTITCU8ciy+C0GNv8B0PY8ZMc/APfRCaC7G30dfZjtb4ZhTir4u8pAANV1B+hvEzBvVj8U8uRIGyhcuw8c+xic3wfO7wenUKI3Yy56ZyxAb/o89GTMQ79hJsABMk4Ax7G/z4wMVr9oNIBsMsYhOI59KWbMkGayJoSQCUQBFEIISTJ6lR79vn44Gh1YNXMVTEYTuNCAyVB5SziOjcjzy1+y6WPH2KgY0QIkOh1r6nD2LBt6gp7ujZg4AtFwnlL7/exmsaODBTza2ljXHkFg+0pNDeZf0GiCiS3jIQZS7t9n+w59tbWNLIlsaDej/n6WZueTT9g9zKOPsq/Z/Pns3kaMy82cOfzjkClOqWRBlMWLWTTwyhXA64XKYIDZPMLkPPHy+VgEU80Bn/8csOEJ4LPPkHrqFOR3moGOfiyd58XtZhU8A3lRuro5XGlMlkQobDx0TvBD4FTBOvtqM4BmAGcBAH1cCvplKsg4IE3vRbqxH5oUH3oEoJdjH4F6HBp0iLsLzc8U+Mmx5SM6Zloa6566dClr3tY4MCSTmH175sxgdGikCZ4GIw6RRgiZ9iiAQgghSUgpVyIjNQMXWy7iftd9ZM3Jgko+cAEfz116djZQU8Puzltbgb//e9a15+GHI7vr6PXsjtrppAygE0wGQAdApwXmagEsZvd3XV0srnX/PtB6K9DbYdjX7+kKIH0esGKMutPcb+Fw5lM5zpyVo+leMCDX0wHU/o69li/14ekNXqxe4cfZD1kQZUJG5+G44E2O+Io2TZKHXA6YTMCCBSyv06VLLIKoULAIoBDW1GQ0A0eK3wGtFpgzJ3iznZoKbNkCAFD19kLoZSMwu1wsOGi3j+6wY03jdWNhz2Us7DiPBR3nsaDnCrRCZH4tldADlY9FPHvbgJttgFIhIF3vRbreC0EuoGuUia9HKlpwheNY6yJZ6J+rHJAPBKa5jg7W9evYMbaRyQQsX84iyosWsYDKlSvjH+QIr1dCx54X+2fK5cF54rQYYZfJgmPYp6ZSt6QkUVZWBgAwmUwoKipKcGnIeLDZbHA6nQCAkpISGAyGEe+LAiiEEJKkOI5Demo6Ovs6cbbpLLIXZAcXajSs/0es3CVqNWsO8PbbbPrmTeBf/xX4z/9k8y0Waad4o5HdrTc1jd8JkbjIAegHXvMB9KtYLpSmCyy4okllNxaJsBjA4jnAV58FbrWk4OTldJy6nIYWT/Dp/NVP2WuWoQ9PZ96Huq4Vjzzkhy7GiEBjIjT/w1A3IyoV+76np7OoVGrqqEajImNAJmOtCubOZQHfzs7gzaZ4kxoeDBvOTaeYP2Ooz1mthloNrLMAN24APgWwcCnwm98Ad+6M7hTHSodyFs6nzsL59BwAANffhzRPIxZ2XsQ87w0s6LuO+b3XoPW5WWOVkG37vRzutSlxr00JbaoPBo1vzO/dlXIBSoUApcIPxTDqKUEABB/gC4mVhQauxDiEmEtKdoOHrJEH9957wUzgy5axDNZj3QJFo2Gj3xmNsR9e+P3B/DP9/cETiPUS85hxHPteinWSTidtekgmTHl5eaKLQMZRQUEBABZIcblcFEAhhJCpTKvSoq27DU2eJszVz2UzDQbg3r3Bk78+9xzrc+FwBDOiulzAb3/LAiuPP866+oh9LEbxz4SMHyWAubOA2SsAlxu4ybP7S7U69n2C+PBzPB9szjcCX3mwB19+tgmXb6Xg2KdGfHJJB2Gga9D9LiX+/WMtfvPJIqw978H2F9uxdNEQmXknSmcnu1H3eoM36KmprJtARkZwWCFqsTKxOI7lt5gx4vF2xqwYDzzAvgqffALs3CkZgTnJqAAsg6xzDtT8XCjal8GvTgUn+AG/H3dalPj4nB6nL+jR2T2B32cBUEJg3Yf0XqTp+pGh90Gn8UEhFyCXCZAP/FTIAJksZJ4cUMgFpKr80KT6oZL5oLp3E8rrl6G8cRmq+zcBIfinq1D4obp8HfJr18e/MUdKCgt2GAzSl9HIloUG/WK1VhHHgheHexP7W967xwIx4roaDRs6bdYs9gWkliqDstvtqKysxObNm1FbW4vMzEwAQH19PdxuNywWy4QGSZxOJ1wuFywWy7gex+12w+FwIDc3F06nEzabDTU1NXC73TCbzdixYwdyc3Nhs9lgtVoBAGvWrMHOnTthNpsD+3E4HONe1tEYTvlsNlsgYDIeKIBCCCGTgDHFiIbmBszQzIBSrmQXbDzPbvpikcuBF14Ann0WOH4c+PDDYObR/n42/eGHLOnsM8+wvuV0gZa0ZDIgPY29urrYtXasJLF+P+sJIQ6bDASf5oa2KB+J8If4HAcsX9iD5Qt70OpuwbFPjTheZ0BXDztAX78MDqcRx+uMWDTfi3SDF0atD2l638DNlQ9pBh/SdV6kGXww6nxIVQuBe5FxoVJFPuH1+1mCmFu3gi1ZMjJYUpf0dHoiPA0ZDGwEZqeT9TBK7niaHpizFjJDJzS3L0HZ3gxwHPRyL764vhtfeuw+PrthgON8BuobDdKmKePFC7jagBtto9uNUiFAl7IKqbrnoNUJyEhzY0nPOSzyNGCe+zzS23koZKzli1olQK0OtlQZcz097HX37tjsT61mgRKtNjhckvi72MVHrw/WRQsWsEAv1UcRcnNzUVdXh9LSUlRUVKC0tDSwzGazwW63x70vnudhtVphs9mwfft2yb4AoKKiAkeOHEFBQcGgXX7cbnfcx3Q4HKisrEReXt6wuhHZbLbA+mazGeXl5bBYLCguLobb7UZubi4A1gLDbrejvLwcJpNJUkaDwYDjx4/DYrEEppPFSMqXmZk5rkEUCqAQQsgkIONkkHNynG8+j6w5WezCKt6O+TodsHkzsHEjcOYM8P770vbo4rAqs2dPgbEupwcNgAfiXFcAiw2IL3H0n5EkkfULQG83iy1Ea8k+G8AqAC+ncbjTokLjXXUgkCIIgH/ga8cNlIsD0AOgCcBdBO/plHIgVe2DNtUPg84Po84PvdYHnUaARu2LvJEd6omv+FOpjH2zIv5uNAYDiT097G9DENiNzPz5rMUWPQ2eNhQKljrK5Rp89PjkoQXwCNDdDa6rE34/i7W3twMPLAFW5QEuN4fTnynR3Dq2ESFBANrdHNraZWhzydDTMzZ/Iz4APX4/ZM094Lz9ADJwQvY5QPYMwAFajQtLus9hiacB6f5mpCj8SFH7oVH7YND7kW7ww6j3I0UlDP/PVhDAdXVA7nFB1ukBfGPciq6nD3D1AWgfuigKFXw6A/yGNGDmTGhXL4Jy5VLWXGqwhymjxIW9aUKMa4/q6mrs3LkzMF1UVITq6uqo665btw5nzpwJTJ86dQrr1q0bdVnT0tIi5jmdThQUFATyX8TDZDKhvLwcNTU1yMrKinqcQ4cOjWmgwWKxIC8vb1jb2Gy2qK0ycnNzYbFY4HA4AsEkq9WK3NxcSfAEYMEisdUKz/PYsmVLIOiSDEZSPrPZDJvNNm7BIAqgEELIJKFT6dDU0YSFhoVI1xnYUym3O/6uNwoF8NhjLMHsxYsskHL+fHD5vXvjUm6SWBxYXpWxehjr97NBTHp6gl2FwqkALAWwNAPo6JahuV0JT488rhQlAd0A2oFeAbgH9hJjISrlCG6E4uTnFOhW6OBRzYRbPYP9HHh55PfhVvrRJ+8AUtRAqgZQ0qXUdGA0JroEw5U68AJmrGS91q5dY2mulBnA018a/xJ0d7P85OKrvZ01ghQDuD5f8BU+r7+ftbTr6goGrnzqFBbQ8PaD6+sF198LThDQyelQr3sC9bonJMcXAPg7Ab+H1YMyeXCUoBERBGj9buh9Lhi8rdD72qH3tcHgbYPe1wal0AeZ4AcHP2TwgxP84CBABjZqkgxsmVLoQ6qvAxp/B7hhNwXqHHjdAodPoVb6kaLys1ROKgEpKj9UCmGUJzq17N27F3v27BlR9528vDzYbDbJDbvYsiEZWmkcP348ZiuLqqoqZGdnY9++fcjKykJdXR327NkTsZ7ZbEZubi7a29uRlZWVVMETYOTlKygoQE1Nzbi0QqH/+oQQMokY1AZ8du8zPGl6EvLsbODCBZYgNj09/jtTjgNWrmSvpibggw+AkyeDfT0IGYSYLkStjh5I8QuA4A82kEpV+bF4bi9kMsDr49DTx6Gvj0O/j4PXx8Hn4+D1c/D52bTXK72lCP9a+3xAt2887w584OCCAS4YcCXqGt0yLdrkM+GSZ6Cfi2/c6hNpY1hEQoZJqwUyM9nANTdusFYpwOBptEZCzNcr1hOpqazh1kgJAqtnurpYEKizk0NXlwqdnSp0denQ2doLd1MXXE1daG2XobM3mPwpEDwWqwvfSHsucQPDA8nQKTeiU25Ek2rRyE8q5ORS/Z1I9XdA6/NA4+8I+d0Drc+NNG8L0rz3ke5thlLoDducQ3efHN19crR1hDS8G6x6nD36Yk8Wbrcbe/fuHdU+CgoKkJ+fH2jJYLfbYbFYxix4IpZRbOXC83xE9xqbzQaz2Qye55GZmRnIW8LzfNQWNyKDwYDy8nKUlZWhuLgYR48ejbkuz/PYs2fPsLo5jQWbzRYYHUfsKmW1WlFdXY2cnByUlpbCZDKNqHxmsxl79+6lAAohhEx3CpkC/n4/rrVfw7KMZcDq1aw/dF0de0Q63KvhuXOBl14CvvQl1qc7mcbrJElNBvZ8W+ljqUPu3x/IjZgKaLTsp1IFqGLEF/p9QF8vuznq7AS6OoG+fhZ88XTJ4OpUor1DAVeXHK4OBVydCrR3KNDRLY8YXYTDwJPewBPf0Ke/A0+EBfbkV+v3QOPrgMbPblg0/g5ofMHflULfkOfObno6Mb//RvxvWFr8qxIyXtRqYMUK4MEHB8+jNFK9vSw/c2jqB5UKI85pxHFs25QUlgYkbCmAFAApbHCbnl703neh9b4Pre0ytLpkaGmXo90lQ2u7DD19Iwu8cn4/OG8vZN4ecIG+jwLAySDI5BBGlRxHBSAD3chAd6xVBkbu0Xg9MPbfh6G/GZruVqg62pDmvY80XwsUQrwf5LVRlHXy8Hg8qKysRHt7+6BBhqGYzWaYTKZAEMNkMkV0gRmNbdu24cCBA4GAjMPhiFh+6NChwPSuXbsCQQWn0ylJAhtNQUFBIEBht9tj5lYRgwwjbX1SUVEx5Dpbt26NeO8KCgqQl5eHjRs3oqGhAQDQ3t6OqqoqSdekkZbv5s2bw1o/XhRAIYSQCXDx4kU8bno8MH3q1qmo6x3610P42//5t4HpF37vBfzVG38lWcegNuBa2zW8uOlFnP3kbGC+459/iPWrV0Opl7Y1P33uHNZ/61uB6bWrVuH0v/6r9MAaDbBkCXb8zd/A+qtfBWbv/cu/xI78/Khl5davl0wLp6KfU/WhQ9j5t8FzKnrhBVT/1V9FXXfd7/8+zoR0Kzr15ptYt3p1xHpxndMAOqcJOqecGOf0UvRzUg68tAAyopxT1Z//JQq35gdieuLon/1eYHau9Jwu/OLTqOW0vfPvKKv+XmD6pY35eP2PyqKum/8/vgHntXOB6d985+/wiCEdCncr5K5WyN1tkLta8dmdG3jaGXwPH9Ho8LvVa6Pu89UbF3GgmQ0NvmLeiqjrTKSLFy9K8hlMZC6D06dPY33Id3Ht2rU4ffp01H3u2LEjMFoEwJrg79ixI+q6iczPQOcU+5wEgXXh6exk3XdWrpSe0zvvRO+G91//VY1/+IfgOW3ZUoQ///PgOYWmNNqxYx0uXgye009/egoPPbQOxmXAcgUL6CoUQEPDaXzpS8FzeuihtTh6NPo5/cVf7MCbbwbPqbJyL/7gD0I+p4E+jFxPN2atlI4W1fbvUZ7wc8CBt/8Lf7b3HwKzvr1pC/5+559L1uEGTu6Z0j/C2auXAove+/t9eGSlGcLAyQhyAyDPwKeXz2HTtsLAeqb5D6Nw6+/QdFcGd4f0jf3401dw9cZ+AMCKWcOvh2J9/8Lt2LEj5nc6XKzv1FjR6/UoLS0Fz/MoK4te58eroKAA1dXVqKqqkgQsxNYj4Qlm4+VwOGA0GiWtWUIDDHa7HQsXLpRss2HDBlitVpSXl4Pn+SEDKE6nE8aB/oeVlZVRc6CMhZG+BwBrKXPgwAHk5+ejsLAQFotlzEYDMhqN45IHhQIohBCSAH925M+izr/jvBN1vs1pw4mbJwLTfsGPyy2XJevsuPUjGHuAL6Y+jK3LX0BGSjo0Sm3Evq503sTzH0c//mf3LkWd/x3n93Glc/BIvrjP3ctfxmPpmTHXO3KvdmBdAX5BgAAhcIF2uZOXrNvZ1wlPrwcvffKXkvmuxg7J9OVOHnknXgUA/CjzfyFDGfxn2dHfKVl3z9WD+NWJ/x70XESeXg+udN3Cdy/8k2Q+f1U6AkPN3ePIO/Eq0pV67M38rnQfXunxX617A0ZX5Jio4efk8/vQ3tOOd5o/xo9v/kayzHn3qmS66urPcejEe3hUvwL/48FvYTC5J16JOj/8nHq9vWjrboOV/zXea5Ve7IZ/Tq/U/R2MLh2+NufzeHHuFwLz3b3uiO1iHT/8nHzyLngVbfjuxX/G9e7ofxeiXXf/CABQ8sDvYZ1xVWC+zih9Tx3dHwbWDcd7GyXTaQ+lIX3VCnz9jPTzdN3oAUJyEd7RefHXT7Ww4z/4e9DLtVDKFFDKlVC++Sbwu6ZBy55IX/zZF6POb/xY+l74BB/6fH3455P/jHeuviNZdrlVWg8V1xTDeN6Il8wvYWvm1sD8Pm9fxHZf/Lfox6+7XCeZ7vf1o9fbi5K3S3C17WrUbQLnNLDP7z75XWQvyJbsI5T9sj3q8TmOw5XW6F23nv/585Lp9qvtkukrrVcC6/z0qz9FRmqwyUSvV9r94ocf/xD/qf/PQc8l1OXWyxH/N258LG0BdeTyETz/8+eRkZqBn371p5JlnWH14Hfs30HaxbSI44Sfk8h+2Y5/OimtBz+7/JlkWjyn7PnZKHu6LDBS76xZkfv7QZP0vRRv+Rtd0nOaN48NEvfPp36It68ekSy72yf9nKz8d5CmTMM3Mr+Bb67+ZmB+a6v02DfcV/CHR6XHD5zTTek56fWs/N+xfwdX2qJ/L0Tf6v4BAAG7Ld/FY/OD3z3NnfOS9f7bW4c/6P0BQiIngTfgul+ai8zw1CNIX7cu8rt3s10y7VZdx5n17JzDv3s7dihw1YppyWQySbpwjORm2mAwIDMzM+Km3uFwRE0wGy+n0wm9Xh9zeV1dXUTrGaPRCH6g/117e3sgOBJLZWUlqqqqYLPZUFlZieLiYkmLlniNNlg0FLPZjO3bt2Pfvn3DGoFIFKt8er0eLpeLAiiEEELYqDzhTykVcgVkRh04wyx03L6Be7o78HECLtyVBj68ghdtPVHGlRSAXp/0Iv9ayxV8dMOBJvdttPW1Rm4Toq2bLa9vqoMQ0n77Wov0orPX14vWLnbDGX4OPkE6zMW1lstQN/XBE3Yj3tUnbezsE3yBdS7cbYBREQwctXdLz7XH2x2xv1gamurA996PWL/HKz1+v78fnl43ZF4fGpqkN399YTdOXX2dkPVGDoETfk493m5cuNuAW55GdPZKAwH9funNYJ+3F529HWiRNePSPenFerjwfcUqp6u7HZfuncd9z92IbcI/p56+bih6gSb3bVySBY/f2Bp5Ix7r+OHndM/dhEv3zqO9qw2dfdG3CT+nm63XYewJzr/vkgZe+n396OzxsInQp6qCAL9fek43+Hpo5O3o7GoPDhkkAN090s+pQwmcWMD2VZfhg0HuRXqqHsszlrHudUmsZeDvMFxnn/Rm+47nDo5dP4bLrZcjtvH6pbmTXL0u+Lp9uNByAR82fhiYf7HpomQ9n9+H1u7odUp4sOFy62V82Pghbntux9xGJC4/e/espD4LD/T0enuj7kuAEPFdPHnrJFwZLjR3NUvmd/RIv5devzewzgc3PoBBHbxov9Mh/S529HVE7C+W9669h1vuW5HHD/u76PH1oLmrGb3eXrx37T3Jsq7+Lsl0e087vF2Rea/Cz8nT58F7197Dp02fRhy/x9cjmRbP6Wrb1Yjjh2vrjX7unV7p8Zs6b+NY43u41Hox4vjh3z3xnJz3nJLjX7x1MWK7WO99+Dmdbz6P9669B97ND/l5Nfey79Opljp0cMHv0HmX9H9gj78XzT3Rv8fh5zSa795c7Vysmb1m0DJPB2K3D4fDAZ7nh50TQxw+N5TT6UR1dTV27Ngx4hYOZrMZNTU1MZcvWrQIx48fl8xzuVyBFiRpaWlwuVwxt6+oqEBRUREMBgOKiopQU1MDp9MJq9U67CDFUMGikXbhEbndbjQ0NASGX3733XeH9Z7GKp/H4xmXFjcUQCGEkClEANCRrkXr4kXQn78Kg3Em5uvmSdaRcTJo5CmA3w9OGLgz5ABBLodCJv23kKGbhcUzlkLfmgZN7B7aAACNigUt5hkXYLFhacg+pBevCpkCqUoNfPDB7/dDEPyxE/vJFeAUSnCctI95+HToPE6hBKcISbwRZdto20fDKZTgfIr4jy/jpMdmS+I6fuQ8DpxcAS6eoSO4gZdMFn1YnNjFiT1fzAwrk438+LIoZYk3DYG4L46L7/hg32G/UhGcVsgj1hPkA+/nQMZFYeAJsBCWebF32RJ0Za6Ar1YPgQs+KfZ5wvYpl8OXzp4Cdq17CHK1HumGhcCM5cCbb8Z5sokR/vcukoV9F1VyFfRqPdQKdcQ24UFQOSeHXCaHWq6GRhkcFj1FkSI9CAfIo30/ouxTJVdBo9RAIVPE3CZw/IHlKYoUyfFVclXEMaLtS8bJIs7/wfQHYZ5lhl4lfVosKKU1l0wmC6yzasYqpKemB5alpaRJ1k1RpETsLxbzLDM0Ck3E+i6F9OZJKVNCr9IjLSUN5lnSpv1quVoyrVVqox4//JxSFakwzzLjbsfdiPWVMmldJ57TbO3siOOHi3Xu4eeUnpIO8ywzTtw8EbGNLCz3iHhOCw0LJcf3pnsjtot1/PBzmq+bD/MsM2akzoCn1xPXOS1JWyI5/hndGcl64ucUTfg5jea7F/E3Nw20t7dHne90OlFZWYnXX3992Pusra2VdJUDWPDDaDSOasQaMSgTGoCpq6sL3PCL+UtC2e32wGhCBoMh0BolnNPphMfjkQR+qqqqsGnTpqhdedxuN2pqamAymZCZmSkJXsQTLBpty5TXXnsNVVVVAID8/PyI3C+jLd9Y44R4O7ZNYitXrsSFCxcSXQxCSIIkQx2wcuVKvPnum0Ne/Mero68DAoSIiz2FTIE52jmYoZkBrZeD7MwnLEugmECC41incIOBZeQzGFgba7U6xpFIIglhXZzI+OA4Dhy4iBv3sZQs9VCiy0AISZypXAfY7XZUVlZi8+bNqK2tRU5ODgCgoaEB9fX1yMnJwfe+972oN9hlZWURwxw7nU4cPnwY+/btQ0lJiWToYrfbjcrKykGHRnY6neB5ftAgC8/zsNvtgVwmdrsdDocD5eXlsFgsgeXiSDRmszkQFOF5PpAPJfQ9OHz4MGpra2EymbBjx47A8a1WK2w2W2CkH4vFgpKSEhgMBlRUVGDnzp2BFi3hrTYKCwuxf//+mOcxUlarFTU1NXC73YFRggoLC+FwOGCxWFBQUIDc3NwRly/afJvNBovFMqqWKdQChRBCJhmf3wcBAp5c9GTMJ8gAADWAnBw2XmVqKhvHUqMZuoUCSRrijX3crTYIIYSQaSg3N3dULULCmc1mmM3mqK0r6uvrA0MLhyeCHQ6TySTpThPeVSh8efiy8BY3g70HRUVFMfe1ZcsWvPbaa0hLS4sICrnd7nHpBhOrTNECISMpn9PpHNPvQ6jRjLtFCCEkAdy9bmTNzho8eCJSKIAlS9hwxXo9BU8IIYQQQkZBvGGfqC4jsWzYsAFOp3PoFQfhcDjgcrmwZ8+eqEltQ4NFbnd8+ePG0kjLZ7PZkJeXNy5lSooAyq5du4b14TudTthsNjgcjkBTJzI5/eM//mOii0CimIqfS6w6xm63w2azoaKiAoWFhbDb7XHtz+FwwGq1wmq1oqKiIqKf6njp6u/CDM0MzNDMGHplEpep+H2fKqbaZzNV6iEy9qbad30qoc+GhBNH9xlqGOHxVlBQgMOHD49qH2IwyOl0YsOGDYH5mzZtkixPVLBoJOVzOp2wWCzjVt6EdeER+23p9XrU1tZi69atQ280sN3evXuxZ8+ewLxdu3bBaDQm/EtMhu+HP/whXn311UQXg4SZCp9LaB0DIGqmcrHfaWjlm5+fD57nB81QLgZtQ9cRb2RGMvxavARBQK+vF4/NemzcjjEdTYXv+1Q12T+bqVgPkfEx2b/rUxl9NmSkJiLgsHPnTtjt9hF3VzGZTFG7wIhdacKHgp5oIylffX39uJY5YS1QTCYTysvLUVpaOuQY1qGsVmtEsGXnzp2orKwc6yISQiax0DomltDh4AD2j66kpGTI+uTgwYMR/VQtFgvq6upibDE22nvasXrmaqgVlPCVkMlgKtZDhBAy1TgcDpSVlcFqtY7ZPkMTvo4ng8EwLrk+EtFdZzgGK1+04InNZkNZWdmYtNScdElka2pqIp6smEwm6sZDCBkWnudRVlaGzMxMSes18Xen0xmzVdvNmzcDGcInSq+3Fzq1DvP18yfsmISQ8TXZ6iFCCJmKxBFgSFCy9+wYbvnGskVKUuRAiZeYHCa8GU9ofydCCImHyWTC9u3bI+oTMaI9WMbxHTt2oLCwUBLFtlqt2Llz5/gUFmzY4qzZWeM6zCohZGJNtnqIEEIIme4mVQuUoZoSRetbLFq5cuVYF4eMEfpsktN0+FyiNas/fPgwzGbzoP1Wc3NzUV5eHmgKmJeXh4KCgiH7un5r47dGXWYyPqbD932ymuqfzUTXQ1P9/ZzM6LNJXvTZEEJEkyqAMlIXLlxIdBEIIQkkZuoeitvtxltvvYVDhw4Nua7FYsH27dtRW1uLyspKGAyGQZsHUj1EyPRG9RAhhBAy+U2LAAohhMSjuLgYVVVVgzabB1iyMafTGXhyLCamcjqdKC8vn4iiEkKmKKqHCCGEkOQ1qXKgiM1SY3XlGc5oPoQQEqqiogJFRUVxJWQ8ePCgJJl1QUEBjh49ipqaGsrFRAgZMaqHCCGEkOQ2qQIoJpMJBoMhIteJGFBJ9mzBhJDkZLPZsGHDhrhuWhwOB7KysiLmm0wm7NixAzzPj0cRCSFTHNVDhBBCSPKbdF14cnJy4HQ6JU1b6+vrIy44nE4n6uvrYTKZ4Ha7YTAYaKi/BLHb7XC5XGhsbERDQwMKCgoixiunzyux3G43KisrI5p9T4fPxeFwRAwhKt58RGtCbzKZcPz48aj7MhgMEdtMh/dwMqB6KPlRPUT10FRH9dDkMJ3rIkLI0JI6gOJ2u7Ft2zbs2LEj8A+mtLQUxcXFkn84Bw8eRElJSWCa53ns3bsXe/bsCczbtWsXjEYjtVKZYHa7HWazOXAx53a7kZ+fD57nA02P6fNKvNdeey1i3nT4XBwOR2Bo9NAntgcPHgzkFQivh8R1nU6n5H1wu91obGyUJHCcDu/hZED10ORA9RDVQ1MZ1UOTx3Stiwgh8UlYAMXtdmPv3r3weDzgeR6VlZXIycmJaL7K87yky47JZEJJSQmsVmvgAmLr1q2SystqtWLr1q2S4+3cuROVlZXYv3//+J8cCXC5XJInYQaDASUlJSguLg5cMNDnlVgOhyPiwh2Y/J/LUHWM2+1GYWFh1G1NJpNkaNHwemjPnj2wWq04fPgw0tLSALDvdvhwpJP9PZwqqB5KflQPRaJ6aGqhemhymKp1ESFkDAlT0Pr164XGxkbJPJfLJaxYsSJBJZqeGhsbhRUrVgj19fWDzqfPK3FcLpdw/Phx4eDBg8Krr74qWUafy+jRe5h4VA8lP6qHxhe9h4lH9dDkQHURISQekyqJbDx4ng80hQ0ljuBDmeknjslkwvbt2yM+CzHprxjhp88rcWpqaqL236XPZfToPUwOVA8lP6qHxg+9h8mB6qHJgeoiQkg8plwAJdYQx6LwEXzI+CotLQ38gxEdPnwYZrMZBoOBPq8EcjgcyMvLi7qMPpfRo/cweVA9lLyoHhpf9B4mD6qHkhvVRYSQeCV1Elky9bjdbrz11ls4dOhQoosyrbnd7kAGeUKmG6qHkgPVQ2Q6o3ooeVBdRAgZjinXAoUkt+LiYlRVVUUdlpFMnJqamoihEwmZLqgeSg5UD5HpjOqh5EF1ESFkOKZcAEWMHsdqbmc0GieyOCRERUUFioqKJP1L6fOaeE6nM2of31D0uYwevYfJieqh5ED10MSg9zA5UT2UPKguIoQM15TrwmMymWAwGOByuSRN8cRKj8ZqTwybzRYxRDVAn1ci8DyPuro6ybza2lq43W5UVFQgKysLubm59LmMEn23kw/VQ8mD6qGJQd/t5EP1UHKhuogQMlxTLoACADk5OXA6nZJmkfX19UNGmMn4cDgcyMzMlPyD4XkeALtgoM9rYuXm5kY0VbVarairq0NpaWlgHn0uo0fvYfKgeii5UD00ceg9TB5UDyUfqosIIcM15brwACzTeXV1tWTewYMHUVJSkqASTV8OhyOQmIvn+cDr4MGDgX9C9HklXnt7e8Q8+lxGj97D5ED10ORA9dD4oPcwOVA9NHlQXUQIGQwnCIKQ6EKMB4fDEYgU8zwPs9lMUeIJ5na7kZ2dHXWZyWTC0aNHA9P0eSWGePF25MgRuFwuvPTSS9iyZUvg6Rh9LqNH72FiUT2U/KgeGn/0HiYW1UOTA9VFhJB4TNkACiGEEEIIIYQQQshYmZJdeAghhBBCCCGEEELGEgVQCCGEEEIIIYQQQoZAARRCCCGEEEIIIYSQIVAAhRBCCCGEEEIIIWQIFEAhhBBCCCGEEEIIGQIFUAghhBBCCCGEEEKGQAEUQgghhBBCCCGEkCFQAIUkNZvNlugiJDV6fwiZGPS3Njh6fwgZf/R3Njh6fwghE4ECKCRpVVRUIC8vDwDgdruxadMmbNq0CStXrgy8Nm3aFPiHKa4jLsvOzsamTZvgcDgSeRrjKi8vDxUVFYkuBiFTGtVFQ6O6iJDxRfXQ0KgeIoRMBE4QBCHRhSAknMPhgNvtRm5ubsSywsLCwAXA/v37YbFYJMvLysoAAOXl5eNf0CRgs9lgMpki3gdCyOhRXRQ/qosIGR9UD8WP6iFCyHijFihkwgznqcfBgwejXigAQFVVFQwGAwCguLgYbrdbcgyHwzFtLhQAoKCgAAcPHkx0MQiZNKguGh9UFxESP6qHxgfVQ4SQ8UYBFDJhQp+SDMbpdCIrKyvmcoPBgNdffx0Aa6L62muvBX4vLi7G/v37x6bAk0hWVhacTmeii0HIpEB10fihuoiQ+FA9NH6oHiKEjCcKoJCkc/jw4ZhPWkS5ubnYvHkzAODIkSOw2+0oLi5GSUkJTCbTRBQzqVgsFkqeRsgYo7po+KguImRsUT00fFQPEULGEwVQpiGHw4Fdu3YhOzs7kHTMbrdHrFdRUYFNmzYhOzsbu3btkjQLFdntduTn5wf2lZ+fj/z8/MBTFbvdHkhgBrAnLqEJz3iej9hnbW1tXP/wv/e970marer1ehQUFAzrvRhLVqtVkrBt06ZNgWW7du2SJHGL9l6OhtlsRn19/Zjuk5DxRnXR+KC6iJD4UT00PqgeIoRMVRRAmWbsdjsKCwthMplQVVWFo0ePwmAwoLq6WrJeYWEhjhw5gvLyclRVVcHj8WDjxo2Sf3J2ux27d+9GQUEBDhw4gEOHDiEzMxNOpzOwXm5uLo4ePYqjR48CYEnMxOmjR4+O6smIwWBAVVVVYPrmzZsj3tdYKCoqwtGjRwNPgULPbevWrTCbzXj33XcD7/lYG+sLEELGE9VF44fqIkLiQ/XQ+KF6iBAyZQlk2nC5XML69euFgwcPRsx3uVyB6YMHDwrr16+XzBMEQXjhhReE3bt3B6bfeOMN4dVXX404zvr164Xjx49HzF+xYkXU+eFl2bZtW1znI66/fv16YcWKFcKKFSuE6urquLcdTy+88IKwYsUK4Y033hAaGxuFjRs3RryfY23btm1CY2PjuB6DkLFAddHEobqIkOioHpo4VA8RQqYSaoEyjYhNSMObdBoMBkn0//jx48jLy4t4IrBjxw7U1NQEprdu3Yra2loUFhbCarUGhtk7efLkiIePc7lc0Ov1ca//2muvYceOHYEnG5WVlVGbwI6E3W6P2ow3HgcOHIDBYMC+fftQWFiI/fv3j/oJy2jKQ0gyobpoeKguImTsUT00PFQPEUIIQwGUaYTn+biahzY0NERdz2AwSJpEmkwmvPvuu7BYLHA4HCguLh60b3A8hrOdzWaDx+NBUVGRpNlqYWHhiI4d7vDhwyNuThvalHasLl5GUx5CkgnVRcNDdREhY4/qoeGheogQQhgKoEwjJpMprmHd1qxZE/UfHM/zkicGDocDDocDRUVF2L9/P06ePImjR4/C4/EEhtEbDM/zEUP4GQwGeDyeuLatrKwM/EM2m80oKSkJLKuoqBhyH0NpaGiA2Wwe8fYHDx4MPHUqLCwcdX/ceMpjNBpHdQxCJgLVRcNDdREhY4/qoeGheogQQhgKoEwj4j8uq9Uqmc/zvOQJxYYNG2Cz2SL+uVmtVuTl5QWmxYRpoUwmE4qKilBbWxu1DKH7tFqtEWUxmUxwuVxDnkthYSGqqqokFy9FRUWBf6b79u2LuBARiRc5Tqcz6sWT0+kMXGxEex/iUVFRAZPJhP3796OgoAA8z6O4uHhcyxN+MUdIsqK6iKG6iJDEoXqIoXqIEEKGhwIo04jBYEB5eTkqKytRUVEBp9MJh8MR0byzoKAAZrM5MPSe0+lEYWEhXC5X4ImGyO12Iz8/H3a7HTzPw263o6ysTHJRITKbzbDZbHA6nbDb7aipqRlRv+CysjJYLJao24Y2Wy0uLo74x+p2uwNPQUwmEw4fPhy1nFlZWdi8eTMKCgok/4DF87XZbDHLZ7Va0dDQgNLSUgAsy77ZbIbD4Yh4CjTa8oSiCwUyWVBdRHURIYlG9RDVQ4QQMiKJzmJLJt7x48eFbdu2CStWrBA2btwovPHGG1HXe+ONN4SNGzcK69evF1599dWIjOnV1dXC7t27hd27dweyvm/cuDFm1vfGxsZAJvb169fHXG/37t0RmdNramoCZRGzy2/cuFGyTn19fdR1Nm7cGMh0L2aof/XVV4WampqY79Grr74aNTt+6PbR3q/QY4vnEJ4Vf+PGjYGs/6MtT+i5J0u2fULiRXUR1UWEJBrVQ1QPEULIcFAAhSSd48ePj/s/vuPHjwsvvPBCzH/A4Rci4cKHPUx0eaqrq4X6+voxLRMh0x3VRcMvD9VFhIwtqoeGXx6qhwgh44m68JCkY7FYUFdXN+b75Xkeu3btChyjoKAgkGAsNEGc2+0ONP2M1Wc42cpTV1c3quRuhJBIVBdRXURIolE9RPUQISS5UACFJKWtW7fCbreP6T5NJhO2bNkS6OdsNBoD/2ArKioCfXENBgMyMzPhcDiiDpEX79CHE1Ueu92OrVu3jro8hJBIVBdRXURIolE9RPUQISR5cIIgCIkuBCHRVFRUYOfOnROWCMzhcMSVwM1qtaKoqCgpyuN2u7F3795AcjZCyNijuojqIkISjeohqocIIcmBWqCQpFVaWoqampoJO168Q/NNxIUCEF95ampq6EKBkHFGdRHVRYQkGtVDVA8RQpIDtUAhBOzJRmZmZtIMe5ds5SGETIxk+9tPtvIQQsZfsv3dJ1t5CCHTGwVQCCGEEEIIIYQQQoZAXXgIIYQQQgghhBBChkABFEIIIYQQQgghhJAhUACFEEIIIYQQQgghZAgUQCGEEEIIIYQQQggZAgVQCCGEEEIIIYQQQobw/wMCLE6RNAzGHAAAAABJRU5ErkJggg==",
      "text/plain": [
       "<Figure size 1080x144 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Each frame corresponds to one time-slice.\n",
    "plot_expected_opt_curve_paper(T,\n",
    "    best_objective_values,\n",
    "    exp_per_trial_cost,\n",
    "    exp_optimal_outcome_values_during_trials,\n",
    "    plot_params, \n",
    "    fig_size = (15,2),\n",
    "    y_lim_list = y_lim_list,\n",
    "    # filename=\"final_50_trials\"\n",
    "    )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "interpreter": {
   "hash": "e54279b2d173705d392bde41516bb8c915f7eee4cde8e160cde6d5088aa4048d"
  },
  "kernelspec": {
   "display_name": "Python 3.9.0 ('bt')",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.0"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}