initial_loss_stats.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "4d1d571c",
   "metadata": {},
   "source": [
    "# Initial loss stats\n",
    "\n",
    "I noticed a weird trend in the last tensorboards: the initial loss of the model is almost always greater on the reversed (i.e. labeled `backward`) dataset than in non-reversed one (labeled `forward`).\n",
    "This looks suspicious, because for randomly initialized weights, the untrained model's loss should be, on average, equal on `forward` and on `backward` datasets.\n",
    "\n",
    "Most likely explanation is, the first point on my tensorboard learning curve plots corresponds to the loss _after the first epoch_, i.e. after some training has already been done.\n",
    "Maybe I should call my tensorboard callback just before the training loop to add the true initial point.\n",
    "\n",
    "Anyway, I explore this trend more rigorously in this notebook."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "81abcbc9",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/conda/lib/python3.8/site-packages/tqdm/auto.py:21: TqdmWarning: IProgress not found. Please update jupyter and ipywidgets. See https://ipywidgets.readthedocs.io/en/stable/user_install.html\n",
      "  from .autonotebook import tqdm as notebook_tqdm\n"
     ]
    }
   ],
   "source": [
    "# Custom code imports\n",
    "from models import ThreeFullyConnectedLayers\n",
    "\n",
    "from generate_time_series import (load_two_body_problem_time_series,\n",
    "                                  load_lorenz_attractor_time_series,\n",
    "                                  load_belousov_zhabotinsky_time_series)\n",
    "\n",
    "from datasets import (prepare_time_series_for_learning,\n",
    "                      time_series_to_dataset,\n",
    "                      AllDataHolder)\n",
    "\n",
    "from train_test_utils import get_mean_loss_on_test_dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "166d97c9",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Standard modules import\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "from typing import Optional"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "bc98910f",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Dataholders\n",
    "def load_two_body_problem_dataholder(window_len: int,\n",
    "                                     target_len: int) -> AllDataHolder:\n",
    "    twb = load_two_body_problem_time_series()\n",
    "    dh = prepare_time_series_for_learning(train_ts=twb,\n",
    "                                          test_ts=twb.copy(),\n",
    "                                          window_len=window_len,\n",
    "                                          target_len=target_len,\n",
    "                                          take_each_nth_chunk=1)\n",
    "    return dh\n",
    "\n",
    "\n",
    "def load_lorenz_attractor_dataholder(window_len: int,\n",
    "                                     target_len: int) -> AllDataHolder:\n",
    "    lrz = load_lorenz_attractor_time_series()\n",
    "    dh = prepare_time_series_for_learning(train_ts=lrz,\n",
    "                                          test_ts=lrz.copy(),\n",
    "                                          window_len=window_len,\n",
    "                                          target_len=target_len,\n",
    "                                          take_each_nth_chunk=1)\n",
    "    return dh\n",
    "\n",
    "\n",
    "def load_belousov_zhabotinsky_dataholder(window_len: int,\n",
    "                                         target_len: int) -> AllDataHolder:\n",
    "    bzh = load_belousov_zhabotinsky_time_series()\n",
    "    dh = prepare_time_series_for_learning(train_ts=bzh,\n",
    "                                          test_ts=bzh.copy(),\n",
    "                                          window_len=window_len,\n",
    "                                          target_len=target_len,\n",
    "                                          take_each_nth_chunk=1)\n",
    "    return dh"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "eafa2d1b",
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_loss_difference(model: ThreeFullyConnectedLayers,\n",
    "                        dh: AllDataHolder) -> float:\n",
    "    loss_forw = get_mean_loss_on_test_dataset(model, dh.forward.test_dataset)\n",
    "    loss_back = get_mean_loss_on_test_dataset(model, dh.backward.test_dataset)\n",
    "    return loss_back - loss_forw\n",
    "\n",
    "def get_loss_difference_kepler(window_len: int = 20,\n",
    "                               target_len: int = 1,\n",
    "                               size: int = 7) -> float:\n",
    "    dh = load_two_body_problem_dataholder(window_len=window_len,\n",
    "                                          target_len=target_len)\n",
    "    m = ThreeFullyConnectedLayers(window_len=window_len, target_len=target_len,\n",
    "                                  hidden_layer1_size=size, hidden_layer2_size=size,\n",
    "                                  datapoint_size=2)\n",
    "    return get_loss_difference(m, dh)\n",
    "\n",
    "def get_loss_difference_lorenz(window_len: int = 20,\n",
    "                               target_len: int = 1,\n",
    "                               size: int = 10) -> float:\n",
    "    dh = load_lorenz_attractor_dataholder(window_len=window_len,\n",
    "                                          target_len=target_len)\n",
    "    m = ThreeFullyConnectedLayers(window_len=window_len, target_len=target_len,\n",
    "                                  hidden_layer1_size=size, hidden_layer2_size=size,\n",
    "                                  datapoint_size=3)\n",
    "    return get_loss_difference(m, dh)\n",
    "\n",
    "def get_loss_difference_belousov_zhabotinsky(window_len: int = 20,\n",
    "                                             target_len: int = 1,\n",
    "                                             size: int = 10) -> float:\n",
    "    dh = load_belousov_zhabotinsky_dataholder(window_len=window_len,\n",
    "                                              target_len=target_len)\n",
    "    m = ThreeFullyConnectedLayers(window_len=window_len, target_len=target_len,\n",
    "                                  hidden_layer1_size=size, hidden_layer2_size=size,\n",
    "                                  datapoint_size=3)\n",
    "    return get_loss_difference(m, dh)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "3c5305a9",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 5.54 s, sys: 1.37 s, total: 6.91 s\n",
      "Wall time: 4.29 s\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "-0.0003549444284172143"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "%%time \n",
    "get_loss_difference_kepler()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "8434c056",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 533 ms, sys: 19.5 ms, total: 552 ms\n",
      "Wall time: 38 ms\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "_ = load_two_body_problem_dataholder(window_len=20, target_len=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "dea138b2",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 62 ms, sys: 4.87 ms, total: 66.9 ms\n",
      "Wall time: 3.13 ms\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "_ = ThreeFullyConnectedLayers(window_len=20, target_len=1,\n",
    "                              hidden_layer1_size=10, hidden_layer2_size=10,\n",
    "                              datapoint_size=3)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "27535957",
   "metadata": {},
   "source": [
    "### As you can see, evaluation of the model takes wastly more time than creating the model and generating the dataset. So, the next cell is going to take a few hours."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "7d5d3040",
   "metadata": {},
   "outputs": [],
   "source": [
    "losses_kepler = [get_loss_difference_kepler() for _ in range(1000)]\n",
    "losses_lorenz = [get_loss_difference_lorenz() for _ in range(1000)]\n",
    "losses_bzh = [get_loss_difference_belousov_zhabotinsky() for _ in range(1000)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "d1ec41a6",
   "metadata": {},
   "outputs": [],
   "source": [
    "def initial_loss_histogram(losses, title_suffix: str = \"\",\n",
    "                           set_xlims: bool = False, bins: Optional[int] = None) -> None:\n",
    "    if bins is not None:\n",
    "        plt.hist(losses, bins=bins)\n",
    "    else:\n",
    "        plt.hist(losses)\n",
    "    \n",
    "    if set_xlims:\n",
    "        plt.xlim(min(0, min(losses_bzh)), max(0, max(losses_bzh)))\n",
    "        \n",
    "    plt.grid()\n",
    "    plt.title(\"Initial loss difference (backward minus forward)\\ndistribution on \" + title_suffix)\n",
    "    plt.xlabel(\"initial loss\")\n",
    "    plt.ylabel(\"count\")\n",
    "    \n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "083838b7",
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "initial_loss_histogram(losses_kepler, \"Kepler\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "5267b29f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAjsAAAHcCAYAAAAwf2v8AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8qNh9FAAAACXBIWXMAAA9hAAAPYQGoP6dpAABcwElEQVR4nO3deVxUVeM/8M+Aw7AOiLImIuKCuD+ohPuCkJJLUa7lmiva45Kp5QJu5FJWZi49hi1SpqWmqYAbbljuu6SGWgriEqCiMDDn94c/7tdxhn1g4Pp5v168dM4999xz5t4ZPtxVIYQQICIiIpIpM1N3gIiIiKgsMewQERGRrDHsEBERkawx7BAREZGsMewQERGRrDHsEBERkawx7BAREZGsMewQERGRrDHsEBERkawx7JAehUKB8PDwItWtVasWhgwZUuxlXLt2DQqFAmvXri2w3r59+6BQKLBv375iL8MUnn8/8uv/d999Bx8fHyiVSjg4OEjlixcvRu3atWFubo5mzZqVS58rMq1Wi0aNGmH+/PlSWXh4OBQKBe7evVuufenYsSMaNWpUrss0hiFDhqBWrVplvpy89VJR5eTk4P3334eHhwfMzMzQu3dvU3epzHTs2BEdO3aUXl+4cAFVqlTBuXPnTNcpE2PYkaG1a9dCoVDg2LFjRmnv8OHDCA8PR1pamlHae9FdunQJQ4YMgbe3N7766iusXr0aABAbG4v3338fbdq0QVRUFBYsWGDinpreDz/8gL///hvjxo0zdVeokvv666+xePFivPHGG/jmm28wceJEU3ep3Pj6+iIkJASzZs0ydVdMpoqpO0AVz+PHj1Glyv9tGocPH0ZERASGDBmisxcCABITE2Fmxsycn/bt2+Px48ewsLCQyvbt2wetVovPPvsMderUkcr37NkDMzMzrFmzRqf+i2zx4sXo168f7O3tTd0VKsSMGTMwbdo0U3cjX3v27MFLL72EpUuXmrorJjF69Gh0794dV69ehbe3t6m7U+74W4r0WFpa6oSdgqhUKiiVyjLuUeVlZmYGS0tLnUCYmpoKAHrBMTU1FVZWVkYNOpmZmUZrq7ydPHkSp0+fRp8+fUzdlQrtyZMn0Gq1pu4GqlSpAktLS1N3I1+pqal6n7nS0Gq1ePLkidHaK45Hjx4Ve57AwEBUrVoV33zzTRn0qOJj2HlBDBkyBLa2trh58yZ69+4NW1tbODk54b333kNubq5O3WfP2QkPD8eUKVMAAF5eXlAoFFAoFLh27RoA/XNU7t+/j/feew+NGzeGra0t1Go1unXrhtOnTxt1PBs2bICfnx+srKxQvXp1vPXWW7h586ZOnZSUFAwdOhQ1atSASqWCm5sbevXqJfUdAI4dO4bg4GBUr14dVlZW8PLywrBhwwpdvhAC8+bNQ40aNWBtbY1OnTrh/PnzevWeP2enVq1amD17NgDAyclJeq8VCgWioqLw6NEj6T1+9nym77//Xhqvo6Mj+vXrh7///ltnWXnnlBw/fhzt27eHtbU1PvjgAwBAVlYWZs+ejTp16kClUsHDwwPvv/8+srKydNpQKBQYN24cNm/ejEaNGkGlUqFhw4bYuXOn3thu3ryJ4cOHw93dHSqVCl5eXhgzZgyys7OlOmlpaZgwYQI8PDygUqlQp04dLFy4sEi/nDdv3gwLCwu0b9/e4PS7d++iT58+UKvVqFatGv773//q/fKJiopC586d4ezsDJVKBV9fX6xYscJgezt27ECHDh1gZ2cHtVqNli1bIjo6usA+xsbGwtraGv3790dOTg5ef/11/Oc//9Gp06NHDygUCvz6669S2e+//w6FQoEdO3YAKPrnJm97+vHHHzFjxgy89NJLsLa2RkZGhvSeNWrUCJaWlmjUqBE2bdpUYP+fVatWLbz66qvYt28fWrRoASsrKzRu3Fjadn/55Rc0btwYlpaW8PPzw8mTJ3XmN3TOTlG3p/zOKzLUZlxcHNq2bQsHBwfY2tqifv360nZuSN75gXv37sX58+elz1feuB49eoTJkydL22j9+vWxZMkSCCEMjmXdunVo2LAhVCoVdu7cif/85z94/fXXdeo2btwYCoUCZ86ckcrWr18PhUKBixcvAgCuX7+OsWPHon79+rCyskK1atXw5ptv6nw/Af93WkJ8fDzGjh0LZ2dn1KhRQ5q+evVqeHt7w8rKCq1atcKBAwcMvg9KpRIdO3bEli1b8n2v5IyHsV4gubm5CA4Ohr+/P5YsWYJdu3bh448/hre3N8aMGWNwntdffx1//vknfvjhByxduhTVq1cH8PQXtSF//fUXNm/ejDfffBNeXl64ffs2Vq1ahQ4dOuDChQtwd3cv9TjWrl2LoUOHomXLloiMjMTt27fx2Wef4dChQzh58qT011toaCjOnz+P8ePHo1atWkhNTUVcXBxu3LghvQ4KCoKTkxOmTZsGBwcHXLt2Db/88kuhfZg1axbmzZuH7t27o3v37jhx4gSCgoJ0ftEb8umnn+Lbb7/Fpk2bsGLFCtja2qJJkyaoU6cOVq9ejT/++AP/+9//AACtW7cGAMyfPx8zZ85Enz598M477+DOnTtYtmwZ2rdvrzNeALh37x66deuGfv364a233oKLiwu0Wi169uyJgwcPYuTIkWjQoAHOnj2LpUuX4s8//8TmzZt1+njw4EH88ssvGDt2LOzs7PD5558jNDQUN27cQLVq1QAAt27dQqtWrZCWloaRI0fCx8cHN2/exMaNG5GZmQkLCwtkZmaiQ4cOuHnzJkaNGoWaNWvi8OHDmD59OpKTk/Hpp58W+F4dPnwYjRo1ynfPYZ8+fVCrVi1ERkbiyJEj+Pzzz/Hvv//i22+/leqsWLECDRs2RM+ePVGlShVs3boVY8eOhVarRVhYmFRv7dq1GDZsGBo2bIjp06fDwcEBJ0+exM6dOzFgwACDy9+2bRveeOMN9O3bF19//TXMzc3Rrl07bNmyBRkZGVCr1RBC4NChQzAzM8OBAwfQs2dPAMCBAwdgZmaGNm3aACj+52bu3LmwsLDAe++9h6ysLFhYWCA2NhahoaHw9fVFZGQk7t27J4X9orpy5QoGDBiAUaNG4a233sKSJUvQo0cPrFy5Eh988AHGjh0LAIiMjESfPn2KdBi7KNtTUZ0/fx6vvvoqmjRpgjlz5kClUuHKlSs4dOhQvvM4OTnhu+++w/z58/Hw4UNERkYCABo0aAAhBHr27Im9e/di+PDhaNasGWJiYjBlyhTcvHlT75DXnj178NNPP2HcuHGoXr06atWqhXbt2uGHH36Q6ty/fx/nz5+X1nmTJk0APF3nTk5OaNCgAQDg6NGjOHz4MPr164caNWrg2rVrWLFiBTp27IgLFy7A2tpaZ9ljx46Fk5MTZs2aJe3ZWbNmDUaNGoXWrVtjwoQJ+Ouvv9CzZ084OjrCw8ND773w8/PT2T5fKIJkJyoqSgAQR48elcoGDx4sAIg5c+bo1G3evLnw8/PTKQMgZs+eLb1evHixACCSkpL0luXp6SkGDx4svX7y5InIzc3VqZOUlCRUKpXOspOSkgQAERUVVeBY9u7dKwCIvXv3CiGEyM7OFs7OzqJRo0bi8ePHUr1t27YJAGLWrFlCCCH+/fdfAUAsXrw437Y3bdqk9z4VRWpqqrCwsBAhISFCq9VK5R988IEAoPN+PN9/IYSYPXu2ACDu3Lmj0+7gwYOFjY2NTtm1a9eEubm5mD9/vk752bNnRZUqVXTKO3ToIACIlStX6tT97rvvhJmZmThw4IBO+cqVKwUAcejQIakMgLCwsBBXrlyRyk6fPi0AiGXLlkllgwYNEmZmZgbfu7z3ZO7cucLGxkb8+eefOtOnTZsmzM3NxY0bN/TmfVaNGjVEaGioXnne+9ezZ0+d8rFjxwoA4vTp01JZZmam3vzBwcGidu3a0uu0tDRhZ2cn/P39dbapZ8cixNP3t2HDhkIIIX7++WehVCrFiBEjdLb3o0ePCgBi+/btQgghzpw5IwCIN998U/j7+0v1evbsKZo3by69LurnJm97ql27tt7YmjVrJtzc3ERaWppUFhsbKwAIT09PvffheZ6engKAOHz4sFQWExMjAAgrKytx/fp1qXzVqlX5btfPKur2NHjwYIN9fL7NpUuXGvzsFMWz6y/P5s2bBQAxb948nfI33nhDKBQKnX4DEGZmZuL8+fM6dTds2CAAiAsXLgghhPj111+FSqUSPXv2FH379pXqNWnSRLz22mvSa0PbZkJCggAgvv32W6ks7/u8bdu2IicnRyrP+y5s1qyZyMrKkspXr14tAIgOHTrotR8dHS0AiN9//93geyRnPIz1ghk9erTO63bt2uGvv/4yWvsqlUr6Sy83Nxf37t2TdjWfOHGi1O0fO3YMqampGDt2rM75ASEhIfDx8cFvv/0GANK5L/v27cO///5rsK28PSLbtm2DRqMpch927dqF7OxsjB8/XmcX+4QJE4o/oEL88ssv0Gq16NOnD+7evSv9uLq6om7duti7d69OfZVKhaFDh+qUbdiwAQ0aNICPj49OG507dwYAvTYCAwN1TmBs0qQJ1Gq1tJ1otVps3rwZPXr0QIsWLfT6nPeebNiwAe3atUPVqlV1lhsYGIjc3Fzs37+/wLHfu3cPVatWzXf6s3tmAGD8+PEAgO3bt0tlVlZW0v/T09Nx9+5ddOjQAX/99RfS09MBPD0s8uDBA0ybNk3vnBNDl1L/8MMP6Nu3L0aNGoVVq1bp7Nlo3rw5bG1tpbEdOHAANWrUwKBBg3DixAlkZmZCCIGDBw+iXbt20nzF/dwMHjxYZ2zJyck4deoUBg8erHMyd9euXeHr65vfW6jH19cXAQEB0mt/f38AQOfOnVGzZk298qJ8dxS2PRVH3md2y5YtRjlPafv27TA3N8e7776rUz558mQIIaTDjHk6dOig937mrcdn13nLli3RtWtX6ZBSWloazp07p7POn11/Go0G9+7dQ506deDg4GBwnY8YMQLm5ubS67zvwtGjR+uc5zdkyJB8T+jP+zyV920bKgKGnReIpaWl3uGnqlWr5hsGSkKr1WLp0qWoW7cuVCoVqlevDicnJ5w5c0b65VIa169fBwDUr19fb5qPj480XaVSYeHChdixYwdcXFzQvn17LFq0CCkpKVL9Dh06IDQ0FBEREahevTp69eqFqKgovfNY8utD3bp1dcqdnJwK/OVcEpcvX4YQAnXr1oWTk5POz8WLF6WTnfO89NJLeic4X758GefPn9ebv169egCg18azv9TyPLud3LlzBxkZGYXec+by5cvYuXOn3nIDAwMNLtcQ8dx5E896/v339vaGmZmZzjkPhw4dQmBgIGxsbODg4AAnJyfp/I687fHq1asAUKR76CQlJeGtt95CaGgoli1bpheGzM3NERAQIP2SO3DgANq1a4e2bdsiNzcXR44cwYULF3D//n2dX3zF/dx4eXnpvM5vmwQMf1by8/y6z/ul+fwhkbzyonx3FLY9FUffvn3Rpk0bvPPOO3BxcUG/fv3w008/lTj4XL9+He7u7rCzs9MpzzvUlPe+5nn+fQcAFxcX1K1bV2+dt2/fHrdu3cJff/2FQ4cOQavV6qzzx48fY9asWdK5QnnrPC0trVTrXKlUonbt2gbHm/d5qsj3QyorPGfnBfLsXwVlZcGCBZg5cyaGDRuGuXPnwtHREWZmZpgwYUK5XzEyYcIE9OjRA5s3b0ZMTAxmzpyJyMhI7NmzB82bN4dCocDGjRtx5MgRbN26FTExMRg2bBg+/vhjHDlyBLa2tuXaX0O0Wq10Iquh9fd8H5/9a/HZNho3boxPPvnE4DKe/0WW33ZSUPAwRKvVomvXrnj//fcNTs8LW/mpVq1asX4hPv8FfvXqVXTp0gU+Pj745JNP4OHhAQsLC2zfvh1Lly4t0fbo5uYGNzc3bN++HceOHTO4Z6tt27aYP38+njx5ggMHDuDDDz+Eg4MDGjVqhAMHDsDFxQUAdH7xFfdzY2g9G0N+674020RR5s3vl+/zF09YWVlh//792Lt3L3777Tfs3LkT69evR+fOnREbG1vm33H5ve9t27bF7t278fjxYxw/fhyzZs1Co0aN4ODggAMHDuDixYuwtbVF8+bNpXnGjx+PqKgoTJgwAQEBAbC3t4dCoUC/fv3KbJ3nfZ7yzr18kTDsUKGK81fAxo0b0alTJ6xZs0anPC0tzSgfME9PTwBP7++TdxgmT2JiojQ9j7e3NyZPnozJkyfj8uXLaNasGT7++GN8//33Up2XX34ZL7/8MubPn4/o6GgMHDgQP/74I955550C+3D58mWdv6Du3Llj1L1kef0XQsDLy6vQcFBQG6dPn0aXLl2M8hedk5MT1Gp1oXdj9fb2xsOHD6U9OcXl4+ODpKSkfKdfvnxZ56/dK1euQKvVSlf1bN26FVlZWfj111919i48f9gu7xDLuXPndO57ZIilpSW2bduGzp0745VXXkF8fDwaNmyoU6ddu3bIzs7GDz/8gJs3b0qhpn379lLYqVevnhR6gNJ/bp7dJp+XmJhY6PymVrVqVYM3LX1+zwrw9HYOXbp0QZcuXfDJJ59gwYIF+PDDD7F3795ib2uenp7YtWsXHjx4oLN359KlS9L0omjXrh2ioqLw448/Ijc3F61bt4aZmRnatm0rhZ3WrVvrhLGNGzdi8ODB+Pjjj6WyJ0+eFPnmrc+u82e/CzUaDZKSktC0aVO9eZKSkmBmZlbi75LKjIexqFA2NjYAUKQPobm5ud5fexs2bNC7LLykWrRoAWdnZ6xcuVLncNOOHTtw8eJFhISEAHh6f5nnL0P29vaGnZ2dNN+///6r19e8RzQUdCgrMDAQSqUSy5Yt05m/sKuLSuL111+Hubk5IiIi9PoqhMC9e/cKbaNPnz64efMmvvrqK71pjx8/LvY9O/Jutb9161aDd+nO62efPn2QkJCAmJgYvTppaWnIyckpcDkBAQE4d+5cvuti+fLlOq+XLVsGAOjWrRuA/9uj8Oz7lp6ejqioKJ35goKCYGdnh8jISL1txtCeC3t7e8TExMDZ2Rldu3aVDoPl8ff3h1KpxMKFC+Ho6CiFoXbt2uHIkSOIj4/X2auT19fSfG7c3NzQrFkzfPPNNzqHQOLi4nDhwoUitWFK3t7eSE9P17lUOzk5We/S+fv37+vNW5TPbH66d++O3NxcfPHFFzrlS5cuhUKhkLalwuStz4ULF6JJkybSYb527dph9+7dOHbsWJHW+bJly/T2ZuWnRYsWcHJywsqVK3WuAl27dm2+39XHjx9Hw4YNX8ibdHLPDhXKz88PAPDhhx+iX79+UCqV6NGjhxSCnvXqq69izpw5GDp0KFq3bo2zZ89i3bp1+R5DLq68XyJDhw5Fhw4d0L9/f+nS81q1akm3gP/zzz/RpUsX9OnTB76+vqhSpQo2bdqE27dvo1+/fgCAb775Bl9++SVee+01eHt748GDB/jqq6+gVqvRvXv3fPuQd3+iyMhIvPrqq+jevTtOnjyJHTt2GH33sLe3N+bNm4fp06fj2rVr6N27N+zs7JCUlIRNmzZh5MiReO+99wps4+2338ZPP/2E0aNHY+/evWjTpg1yc3Nx6dIl/PTTT4iJiTF4OKYgCxYsQGxsLDp06CBdzp6cnIwNGzbg4MGDcHBwwJQpU/Drr7/i1VdfxZAhQ+Dn54dHjx7h7Nmz2LhxI65du1bg+9WrVy/MnTsX8fHxCAoK0puelJSEnj174pVXXkFCQgK+//57DBgwQPqLNigoCBYWFujRowdGjRqFhw8f4quvvoKzszOSk5OldtRqNZYuXYp33nkHLVu2xIABA1C1alWcPn0amZmZBm/CVr16del+L4GBgTh48CBeeuklAIC1tTX8/Pxw5MgR6R47wNM9O48ePcKjR4/0fvEZ43MTGRmJkJAQtG3bFsOGDcP9+/exbNkyNGzYEA8fPixyO6bQr18/TJ06Fa+99hreffddZGZmYsWKFahXr57Oybpz5szB/v37ERISAk9PT6SmpuLLL79EjRo10LZt22Ivt0ePHujUqRM+/PBDXLt2DU2bNkVsbCy2bNmCCRMmFPlOw3Xq1IGrqysSExOlE+WBp+t86tSpAGBwnX/33Xewt7eHr68vEhISsGvXriJfjq9UKjFv3jyMGjUKnTt3Rt++fZGUlISoqCiD241Go5Hu1fNCKuerv6gc5Hfp+fOXNQuR/+Wiz156LsTTy4hfeuklYWZmpnMZuqFLzydPnizc3NyElZWVaNOmjUhISBAdOnTQuRSypJee51m/fr1o3ry5UKlUwtHRUQwcOFD8888/0vS7d++KsLAw4ePjI2xsbIS9vb3w9/cXP/30k1TnxIkTon///qJmzZpCpVIJZ2dn8eqrr4pjx44V2CchhMjNzRURERHSODt27CjOnTun936U9tLzPD///LNo27atsLGxETY2NsLHx0eEhYWJxMREqY6hS2vzZGdni4ULF4qGDRsKlUolqlatKvz8/ERERIRIT0+X6gEQYWFhevM/Py4hhLh+/boYNGiQcHJyEiqVStSuXVuEhYXpXAb74MEDMX36dFGnTh1hYWEhqlevLlq3bi2WLFkisrOzDfb1WU2aNBHDhw/XKct7/y5cuCDeeOMNYWdnJ6pWrSrGjRund+n4r7/+Kpo0aSIsLS1FrVq1xMKFC8XXX39t8FYKv/76q2jdurWwsrISarVatGrVSvzwww/SdEPv75UrV4Sbm5to0KCBzvqcMmWKACAWLlyoU79OnToCgLh69apOeVE/N3nb04YNGwy+Xz///LNo0KCBUKlUwtfXV/zyyy/5Xtb9PE9PTxESEqJXbmibyPv8Pntrh/y+S4q6PcXGxopGjRoJCwsLUb9+ffH999/rtbl7927Rq1cv4e7uLiwsLIS7u7vo37+/3u0NDMnv8/HgwQMxceJE4e7uLpRKpahbt65YvHixzm0HChpLnjfffFMAEOvXr5fKsrOzhbW1tbCwsNDbNv/9918xdOhQUb16dWFrayuCg4PFpUuX9N4bQ9/nz/ryyy+Fl5eXUKlUokWLFmL//v16240QQuzYsUMAEJcvX853DHKmEKKYZx0SEZWT7777DmFhYbhx44ZRb/VP9KLp3bs3FApFse6qLScMO0RUYWm1WjRp0gT9+/fHhx9+aOruEFVKFy9eROPGjXHq1Kki3WJBjhh2iIiISNZ4NRYRERHJGsMOERERyRrDDhEREckaww4RERHJGsMOUQUUHh6u92iHWrVqYciQIWW+7GvXrkGhUGDt2rVS2ZAhQ8r1WWEKhQLh4eHltjwikjeGHSIZ2759e4UNDRW5b2WtvMMj0YuOj4sgqiQSExNhZla8v0+2b9+O5cuXFytUeHp64vHjx1AqlcXsYfEU1LfHjx+jShV+PRGRcXDPDlEloVKpyjSA5OTkIDs7GwqFApaWljpPaC5vlpaWDDulUNyHuxLJHcMOkYkdPHgQLVu2hKWlJby9vbFq1SqD9Z4/Z0ej0SAiIgJ169aFpaUlqlWrhrZt2yIuLg7A00MleU8GVygU0g/wf+flLFmyBJ9++im8vb2hUqlw4cIFg+fs5Pnrr78QHBwMGxsbuLu7Y86cOTpPbt63bx8UCgX27dunM9/zbRbUt7yy5/f4nDx5Et26dYNarYatrS26dOmCI0eO6NRZu3YtFAoFDh06hEmTJsHJyQk2NjZ47bXXcOfOHcMr4Dl79uxBu3btYGNjAwcHB/Tq1QsXL17UqZN3TtWVK1cwZMgQODg4wN7eHkOHDkVmZmaRllMUGzZsgJ+fH6ysrFC9enW89dZbek9CzzskdvXqVXTv3h12dnYYOHAggKd3oP7000/RsGFDWFpawsXFBaNGjcK///6r00atWrXw6quv4uDBg2jVqhUsLS1Ru3ZtfPvttzr1nl1Xz/9cu3bNaOMmMjb+6URkQmfPnkVQUBCcnJwQHh6OnJwczJ49Gy4uLoXOGx4ejsjISLzzzjto1aoVMjIycOzYMZw4cQJdu3bFqFGjcOvWLcTFxeG7774z2EZUVBSePHmCkSNHQqVSwdHREVqt1mDd3NxcvPLKK3j55ZexaNEi7Ny5E7Nnz0ZOTg7mzJlTrHEXpW/POn/+PNq1awe1Wo33338fSqUSq1atQseOHREfHw9/f3+d+uPHj0fVqlUxe/ZsXLt2DZ9++inGjRuH9evXF7icXbt2oVu3bqhduzbCw8Px+PFjLFu2DG3atMGJEydQq1Ytnfp9+vSBl5cXIiMjceLECfzvf/+Ds7MzFi5cWKz3w5C1a9di6NChaNmyJSIjI3H79m189tlnOHToEE6ePKnzrLCcnBwEBwejbdu2WLJkCaytrQE8fZ/z2nn33XeRlJSEL774AidPnsShQ4d09hReuXIFb7zxBoYPH47Bgwfj66+/lp5W37BhQwAwuK5mzJiB1NRUnoNEFZsJH0JK9MLr3bu3sLS0FNevX5fKLly4IMzNzfWeIP3805CbNm1q8CnVzwoLC9NrR4j/e2q1Wq0WqampBqc9+0T6wYMHCwBi/PjxUplWqxUhISHCwsJCeuJ3fk+pN9Rmfn0T4ukTpmfPni297t27t7CwsNB5WvitW7eEnZ2daN++vVSW94TowMBAnadWT5w4UZibm4u0tDSDy8vTrFkz4ezsLO7duyeVnT59WpiZmYlBgwZJZXlP4x42bJjO/K+99pqoVq1agcsQouAn3Avx9GnZzs7OolGjRjpPy962bZsAIGbNmqXTFgAxbdo0nTYOHDggAIh169bplO/cuVOv3NPTUwAQ+/fvl8pSU1OFSqUSkydPzrefixYtEgDEt99+W+iYiUyJh7GITCQ3NxcxMTHo3bs3atasKZU3aNAAwcHBhc7v4OCA8+fP4/LlyyXuQ2hoKJycnIpcf9y4cdL/FQoFxo0bh+zsbOzatavEfShMbm4uYmNj0bt3b9SuXVsqd3Nzw4ABA3Dw4EFkZGTozDNy5Eidw2Lt2rVDbm4url+/nu9ykpOTcerUKQwZMgSOjo5SeZMmTdC1a1ds375db57Ro0frvG7Xrh3u3bun15/iOnbsGFJTUzF27FhYWlpK5SEhIfDx8cFvv/2mN8+YMWN0Xm/YsAH29vbo2rUr7t69K/34+fnB1tYWe/fu1anv6+uLdu3aSa+dnJxQv359/PXXXwb7uHfvXkyfPh3jx4/H22+/XZrhEpU5hh0iE7lz5w4eP36MunXr6k2rX79+ofPPmTMHaWlpqFevHho3bowpU6bgzJkzxeqDl5dXkeuamZnphA0AqFevHgCU6fkad+7cQWZmpsH3pEGDBtBqtfj77791yp8NjwBQtWpVANA7V+VZeUEov+XcvXtX78TfkiynKArqi4+Pj15oq1KlCmrUqKFTdvnyZaSnp8PZ2RlOTk46Pw8fPkRqamqBY8kbj6Gx/PPPP+jbty/atGmDTz75pNjjIypvPGeHqJJq3749rl69ii1btiA2Nhb/+9//sHTpUqxcuRLvvPNOkdqwsrIyap+evxFintzcXKMupzD5XUkmnjmZujItpzAqlUrvtgRarRbOzs5Yt26dwXme36NX1LFkZ2fjjTfegEqlwk8//cSr5qhS4FZKZCJOTk6wsrIyeBgqMTGxSG04Ojpi6NChGDp0KB4+fIj27dsjPDxcCjv5hY+S0Gq1+Ouvv6S9OQDw559/AoB04m7eno20tDSdeQ0dPipq35ycnGBtbW3wPbl06RLMzMzg4eFRpLYK4unpCcDwe3/p0iVUr14dNjY2pV5OcfvSuXNnnWmJiYnS9IJ4e3tj165daNOmjVFD7bvvvotTp05h//79RTqRnqgi4GEsIhMxNzdHcHAwNm/ejBs3bkjlFy9eRExMTKHz37t3T+e1ra0t6tSpg6ysLKks75fz8+GjpL744gvp/0IIfPHFF1AqlejSpQuAp7+kzc3NsX//fp35vvzyS722ito3c3NzBAUFYcuWLTqHy27fvo3o6Gi0bdsWarW6hCP6P25ubmjWrBm++eYbnT6dO3cOsbGx6N69e6mXUVQtWrSAs7MzVq5cqbM+d+zYgYsXLyIkJKTQNvr06YPc3FzMnTtXb1pOTk6JtomoqCisWrUKy5cvR6tWrYo9P5GpcM8OkQlFRERg586daNeuHcaOHYucnBwsW7YMDRs2LPT8G19fX3Ts2BF+fn5wdHTEsWPHsHHjRp2TiP38/AA8/Ws8ODgY5ubm6NevX4n6amlpiZ07d2Lw4MHw9/fHjh078Ntvv+GDDz6QDonY29vjzTffxLJly6BQKODt7Y1t27bpnR9S3L7NmzcPcXFxaNu2LcaOHYsqVapg1apVyMrKwqJFi0o0HkMWL16Mbt26ISAgAMOHD5cuPbe3tzf6oy00Gg3mzZunV+7o6IixY8di4cKFGDp0KDp06ID+/ftLl57XqlULEydOLLT9Dh06YNSoUYiMjMSpU6cQFBQEpVKJy5cvY8OGDfjss8/wxhtvFLm/d+/exdixY+Hr6wuVSoXvv/9eZ/prr71Wbnu+iIrNtBeDEVF8fLzw8/MTFhYWonbt2mLlypXSpc3Pev7S83nz5olWrVoJBwcHYWVlJXx8fMT8+fNFdna2VCcnJ0eMHz9eODk5CYVCIbWZdyn44sWL9fqT36XnNjY24urVqyIoKEhYW1sLFxcXMXv2bJGbm6sz/507d0RoaKiwtrYWVatWFaNGjRLnzp3TazO/vgmhf+m5EEKcOHFCBAcHC1tbW2FtbS06deokDh8+rFMn79Lzo0eP6pTnd0m8Ibt27RJt2rQRVlZWQq1Wix49eogLFy7o1MlbP3mX3D+//KSkpAKXkXe5uKEfb29vqd769etF8+bNhUqlEo6OjmLgwIHin3/+0WuroMvYV69eLfz8/ISVlZWws7MTjRs3Fu+//764deuWVMfT09PgbQw6dOggOnToIIT4v+0iv5/CxkxkSgohyvlMOiIiIqJyxHN2iIiISNYYdoiIiEjWGHaIiIhI1hh2iIiISNYYdoiIiEjWGHaIiIhI1nhTQTy9Df6tW7dgZ2dn1NvrExERUdkRQuDBgwdwd3fXez7csxh2ANy6dcsoz9YhIiKi8vf333+jRo0a+U5n2AFgZ2cH4OmbZYxn7Gg0GsTGxkq3Z3+RcOwcO8f+4uDYOXZTjz0jIwMeHh7S7/H8MOzg/56+rFarjRZ2rK2toVarTb4hlDeOnWPn2F8cHDvHXlHGXtgpKDxBmYiIiGSNYYeIiIhkjWGHiIiIZI1hh4iIiGSNYYeIiIhkjWGHiIiIZI1hh4iIiGSNYYeIiIhkjWGHiIiIZI1hh4iIiGSNYYeIiIhkjWGHiIiIZI1hh4iIiGSNYYeIiIhkjWGHiIiIZK2KqTtARJVLrWm/6ZWpzAUWtQIahccgK1dhgl4V7NpHIabuAhGZEPfsEBERkawx7BAREZGsMewQERGRrDHsEBERkawx7BAREZGsMewQERGRrDHsEBERkayZNOysWLECTZo0gVqthlqtRkBAAHbs2CFN79ixIxQKhc7P6NGjddq4ceMGQkJCYG1tDWdnZ0yZMgU5OTnlPRQiIiKqoEx6U8EaNWrgo48+Qt26dSGEwDfffINevXrh5MmTaNiwIQBgxIgRmDNnjjSPtbW19P/c3FyEhITA1dUVhw8fRnJyMgYNGgSlUokFCxaU+3iIiIio4jFp2OnRo4fO6/nz52PFihU4cuSIFHasra3h6upqcP7Y2FhcuHABu3btgouLC5o1a4a5c+di6tSpCA8Ph4WFRZmPgYiIiCq2CvO4iNzcXGzYsAGPHj1CQECAVL5u3Tp8//33cHV1RY8ePTBz5kxp705CQgIaN24MFxcXqX5wcDDGjBmD8+fPo3nz5gaXlZWVhaysLOl1RkYGAECj0UCj0ZR6LHltGKOtyoZjl//YVeZCv8xM6Pxb0ZTlOnlR1rshHDvHbmpF7YNCCGHSb6ezZ88iICAAT548ga2tLaKjo9G9e3cAwOrVq+Hp6Ql3d3ecOXMGU6dORatWrfDLL78AAEaOHInr168jJiZGai8zMxM2NjbYvn07unXrZnCZ4eHhiIiI0CuPjo7WOUxGREREFVdmZiYGDBiA9PR0qNXqfOuZfM9O/fr1cerUKaSnp2Pjxo0YPHgw4uPj4evri5EjR0r1GjduDDc3N3Tp0gVXr16Ft7d3iZc5ffp0TJo0SXqdkZEBDw8PBAUFFfhmFZVGo0FcXBy6du0KpVJZ6vYqE45d/mNvFB6jV6YyE5jbQouZx8yQpa14DwI9Fx5cZm2/KOvdEI6dYzf12POOzBTG5GHHwsICderUAQD4+fnh6NGj+Oyzz7Bq1Sq9uv7+/gCAK1euwNvbG66urvjjjz906ty+fRsA8j3PBwBUKhVUKpVeuVKpNOqKM3Z7lQnHLt+xF/RU8yytokI+9bw81ofc13tBOHaO3ZR9KIoKd58drVarcz7Ns06dOgUAcHNzAwAEBATg7NmzSE1NlerExcVBrVbD19e3zPtKREREFZ9J9+xMnz4d3bp1Q82aNfHgwQNER0dj3759iImJwdWrV6Xzd6pVq4YzZ85g4sSJaN++PZo0aQIACAoKgq+vL95++20sWrQIKSkpmDFjBsLCwgzuuSEiIqIXj0nDTmpqKgYNGoTk5GTY29ujSZMmiImJQdeuXfH3339j165d+PTTT/Ho0SN4eHggNDQUM2bMkOY3NzfHtm3bMGbMGAQEBMDGxgaDBw/WuS8PERERvdhMGnbWrFmT7zQPDw/Ex8cX2oanpye2b99uzG4RERGRjFS4c3aIiIiIjIlhh4iIiGSNYYeIiIhkjWGHiIiIZI1hh4iIiGSNYYeIiIhkjWGHiIiIZI1hh4iIiGSNYYeIiIhkjWGHiIiIZI1hh4iIiGSNYYeIiIhkjWGHiIiIZI1hh4iIiGSNYYeIiIhkjWGHiIiIZI1hh4iIiGSNYYeIiIhkjWGHiIiIZI1hh4iIiGSNYYeIiIhkjWGHiIiIZI1hh4iIiGSNYYeIiIhkjWGHiIiIZI1hh4iIiGSNYYeIiIhkjWGHiIiIZI1hh4iIiGSNYYeIiIhkjWGHiIiIZI1hh4iIiGSNYYeIiIhkjWGHiIiIZI1hh4iIiGSNYYeIiIhkjWGHiIiIZM2kYWfFihVo0qQJ1Go11Go1AgICsGPHDmn6kydPEBYWhmrVqsHW1hahoaG4ffu2Ths3btxASEgIrK2t4ezsjClTpiAnJ6e8h0JEREQVlEnDTo0aNfDRRx/h+PHjOHbsGDp37oxevXrh/PnzAICJEydi69at2LBhA+Lj43Hr1i28/vrr0vy5ubkICQlBdnY2Dh8+jG+++QZr167FrFmzTDUkIiIiqmCqmHLhPXr00Hk9f/58rFixAkeOHEGNGjWwZs0aREdHo3PnzgCAqKgoNGjQAEeOHMHLL7+M2NhYXLhwAbt27YKLiwuaNWuGuXPnYurUqQgPD4eFhYUphkVEREQVSIU5Zyc3Nxc//vgjHj16hICAABw/fhwajQaBgYFSHR8fH9SsWRMJCQkAgISEBDRu3BguLi5SneDgYGRkZEh7h4iIiOjFZtI9OwBw9uxZBAQE4MmTJ7C1tcWmTZvg6+uLU6dOwcLCAg4ODjr1XVxckJKSAgBISUnRCTp50/Om5ScrKwtZWVnS64yMDACARqOBRqMp9Zjy2jBGW5UNxy7/savMhX6ZmdD5t6Ipy3Xyoqx3Qzh2jt3UitoHk4ed+vXr49SpU0hPT8fGjRsxePBgxMfHl+kyIyMjERERoVceGxsLa2troy0nLi7OaG1VNhy7fC1qlf+0uS205deRYti+fXuZL0Pu670gHPuLqSKMPTMzs0j1TB52LCwsUKdOHQCAn58fjh49is8++wx9+/ZFdnY20tLSdPbu3L59G66urgAAV1dX/PHHHzrt5V2tlVfHkOnTp2PSpEnS64yMDHh4eCAoKAhqtbrUY9JoNIiLi0PXrl2hVCpL3V5lwrHLf+yNwmP0ylRmAnNbaDHzmBmytAoT9Kpg58KDy6ztF2W9G8Kxc+ymHnvekZnCmDzsPE+r1SIrKwt+fn5QKpXYvXs3QkNDAQCJiYm4ceMGAgICAAABAQGYP38+UlNT4ezsDOBp0lSr1fD19c13GSqVCiqVSq9cqVQadcUZu73KhGOX79izcvMPM1laRYHTTaU81ofc13tBOHaO3ZR9KAqThp3p06ejW7duqFmzJh48eIDo6Gjs27cPMTExsLe3x/DhwzFp0iQ4OjpCrVZj/PjxCAgIwMsvvwwACAoKgq+vL95++20sWrQIKSkpmDFjBsLCwgyGGSIiInrxmDTspKamYtCgQUhOToa9vT2aNGmCmJgYdO3aFQCwdOlSmJmZITQ0FFlZWQgODsaXX34pzW9ubo5t27ZhzJgxCAgIgI2NDQYPHow5c+aYakhERERUwZg07KxZs6bA6ZaWlli+fDmWL1+ebx1PT89yOfmQiIiIKqcKc58dIiIiorLAsENERESyxrBDREREssawQ0RERLLGsENERESyxrBDREREssawQ0RERLLGsENERESyxrBDREREssawQ0RERLLGsENERESyxrBDREREssawQ0RERLLGsENERESyxrBDREREssawQ0RERLLGsENERESyxrBDREREssawQ0RERLJWxdQdIHpR1Zr2m6m7QET0QuCeHSIiIpI1hh0iIiKSNYYdIiIikjWGHSIiIpI1hh0iIiKSNYYdIiIikjWGHSIiIpI1hh0iIiKSNYYdIiIikjWGHSIiIpI1hh0iIiKSNYYdIiIikjWGHSIiIpI1hh0iIiKSNYYdIiIikjWGHSIiIpI1hh0iIiKSNYYdIiIikjWThp3IyEi0bNkSdnZ2cHZ2Ru/evZGYmKhTp2PHjlAoFDo/o0eP1qlz48YNhISEwNraGs7OzpgyZQpycnLKcyhERERUQVUx5cLj4+MRFhaGli1bIicnBx988AGCgoJw4cIF2NjYSPVGjBiBOXPmSK+tra2l/+fm5iIkJASurq44fPgwkpOTMWjQICiVSixYsKBcx0NEREQVj0nDzs6dO3Ver127Fs7Ozjh+/Djat28vlVtbW8PV1dVgG7Gxsbhw4QJ27doFFxcXNGvWDHPnzsXUqVMRHh4OCwuLMh0DERERVWwmDTvPS09PBwA4OjrqlK9btw7ff/89XF1d0aNHD8ycOVPau5OQkIDGjRvDxcVFqh8cHIwxY8bg/PnzaN68ud5ysrKykJWVJb3OyMgAAGg0Gmg0mlKPI68NY7RV2XDsRR+7ylyUZXfKlcpM6Pxb0ZTl9shtnmN/0VSksRe1DwohRIX4dtJqtejZsyfS0tJw8OBBqXz16tXw9PSEu7s7zpw5g6lTp6JVq1b45ZdfAAAjR47E9evXERMTI82TmZkJGxsbbN++Hd26ddNbVnh4OCIiIvTKo6OjdQ6RERERUcWVmZmJAQMGID09HWq1Ot96FWbPTlhYGM6dO6cTdICnYSZP48aN4ebmhi5duuDq1avw9vYu0bKmT5+OSZMmSa8zMjLg4eGBoKCgAt+sotJoNIiLi0PXrl2hVCpL3V5lwrEXfeyNwmMKrVNZqMwE5rbQYuYxM2RpFabujp5z4cFl1ja3eY6dYzedvCMzhakQYWfcuHHYtm0b9u/fjxo1ahRY19/fHwBw5coVeHt7w9XVFX/88YdOndu3bwNAvuf5qFQqqFQqvXKlUmnUFWfs9ioTjr3wsWflVrxQUFpZWkWFHFd5bIvc5jn2F01FGHtRl2/SS8+FEBg3bhw2bdqEPXv2wMvLq9B5Tp06BQBwc3MDAAQEBODs2bNITU2V6sTFxUGtVsPX17dM+k1ERESVh0n37ISFhSE6OhpbtmyBnZ0dUlJSAAD29vawsrLC1atXER0dje7du6NatWo4c+YMJk6ciPbt26NJkyYAgKCgIPj6+uLtt9/GokWLkJKSghkzZiAsLMzg3hsiIiJ6sZh0z86KFSuQnp6Ojh07ws3NTfpZv349AMDCwgK7du1CUFAQfHx8MHnyZISGhmLr1q1SG+bm5ti2bRvMzc0REBCAt956C4MGDdK5Lw8RERG9uEy6Z6ewC8E8PDwQHx9faDuenp7Yvn27sbpFREREMsJnYxEREZGsMewQERGRrDHsEBERkawx7BAREZGsMewQERGRrDHsEBERkawx7BAREZGsMewQERGRrDHsEBERkawx7BAREZGsMewQERGRrDHsEBERkawx7BAREZGsMewQERGRrDHsEBERkawx7BAREZGsMewQERGRrDHsEBERkawx7BAREZGsMewQERGRrDHsEBERkawx7BAREZGsMewQERGRrDHsEBERkawx7BAREZGsMewQERGRrDHsEBERkawx7BAREZGsMewQERGRrDHsEBERkawx7BAREZGsMewQERGRrDHsEBERkayVKOx07twZaWlpeuUZGRno3LlzaftEREREZDQlCjv79u1Ddna2XvmTJ09w4MCBUneKiIiIyFiqFKfymTNnpP9fuHABKSkp0uvc3Fzs3LkTL730kvF6R0RERFRKxQo7zZo1g0KhgEKhMHi4ysrKCsuWLTNa54iIiIhKq1iHsZKSknD16lUIIfDHH38gKSlJ+rl58yYyMjIwbNiwIrcXGRmJli1bws7ODs7OzujduzcSExN16jx58gRhYWGoVq0abG1tERoaitu3b+vUuXHjBkJCQmBtbQ1nZ2dMmTIFOTk5xRkaERERyVSx9ux4enoCALRarVEWHh8fj7CwMLRs2RI5OTn44IMPEBQUhAsXLsDGxgYAMHHiRPz222/YsGED7O3tMW7cOLz++us4dOgQgKeHz0JCQuDq6orDhw8jOTkZgwYNglKpxIIFC4zSTyIiIqq8ihV2nnX58mXs3bsXqampeuFn1qxZRWpj586dOq/Xrl0LZ2dnHD9+HO3bt0d6ejrWrFmD6Oho6bBZVFQUGjRogCNHjuDll19GbGwsLly4gF27dsHFxQXNmjXD3LlzMXXqVISHh8PCwqKkQyQiIiIZKFHY+eqrrzBmzBhUr14drq6uUCgU0jSFQlHksPO89PR0AICjoyMA4Pjx49BoNAgMDJTq+Pj4oGbNmkhISMDLL7+MhIQENG7cGC4uLlKd4OBgjBkzBufPn0fz5s1L1BciIiKShxKFnXnz5mH+/PmYOnWq0Tqi1WoxYcIEtGnTBo0aNQIApKSkwMLCAg4ODjp1XVxcpCvBUlJSdIJO3vS8aYZkZWUhKytLep2RkQEA0Gg00Gg0pR5LXhvGaKuy4diLPnaVuSjL7pQrlZnQ+beiKcvtkds8x/6iqUhjL2ofShR2/v33X7z55pslmTVfYWFhOHfuHA4ePGjUdg2JjIxERESEXnlsbCysra2Ntpy4uDijtVXZcOyFW9SqjDtiAnNbGOd8PmPbvn17mS+D2/yLiWM3rczMzCLVK1HYefPNNxEbG4vRo0eXZHY948aNw7Zt27B//37UqFFDKnd1dUV2djbS0tJ09u7cvn0brq6uUp0//vhDp728q7Xy6jxv+vTpmDRpkvQ6IyMDHh4eCAoKglqtLvV4NBoN4uLi0LVrVyiVylK3V5lw7EUfe6PwmHLoVflQmQnMbaHFzGNmyNIqCp+hnJ0LDy6ztrnNc+wcu+nkHZkpTInCTp06dTBz5kwcOXIEjRs31hvsu+++W6R2hBAYP348Nm3ahH379sHLy0tnup+fH5RKJXbv3o3Q0FAAQGJiIm7cuIGAgAAAQEBAAObPn4/U1FQ4OzsDeJo21Wo1fH19DS5XpVJBpVLplSuVSqOuOGO3V5lw7IWPPSu34oWC0srSKirkuMpjW+Q2z7G/aCrC2Iu6/BKFndWrV8PW1hbx8fGIj4/XmaZQKIocdsLCwhAdHY0tW7bAzs5OOsfG3t4eVlZWsLe3x/DhwzFp0iQ4OjpCrVZj/PjxCAgIwMsvvwwACAoKgq+vL95++20sWrQIKSkpmDFjBsLCwgwGGiIiInqxlCjsJCUlGWXhK1asAAB07NhRpzwqKgpDhgwBACxduhRmZmYIDQ1FVlYWgoOD8eWXX0p1zc3NsW3bNowZMwYBAQGwsbHB4MGDMWfOHKP0kYiIiCq3Et9nxxiEKPzKDUtLSyxfvhzLly/Pt46np2e5nIBIRERElU+Jwk5hj4T4+uuvS9QZIiIiImMr8aXnz9JoNDh37hzS0tIMPiCUiIiIyFRKFHY2bdqkV6bVajFmzBh4e3uXulNERERExlKsp54X2JCZGSZNmoSlS5caq0kiIiKiUjNa2AGAq1evIicnx5hNEhEREZVKiQ5jPXv3YeDpVVXJycn47bffMHjwYKN0jIiIiMgYShR2Tp48qfPazMwMTk5O+Pjjjwu9UouIqLzVmvZbmbWtMhdY1Orp4z+Meffoax+FGK0tohddicLO3r17jd0PIiIiojJRqpsK3rlzB4mJiQCA+vXrw8nJySidIiIiIjKWEp2g/OjRIwwbNgxubm5o37492rdvD3d3dwwfPrzIj1snIiIiKg8lCjuTJk1CfHw8tm7dirS0NKSlpWHLli2Ij4/H5MmTjd1HIiIiohIr0WGsn3/+GRs3btR5gGf37t1hZWWFPn36SA/4JCIiIjK1Eu3ZyczMhIuLi165s7MzD2MRERFRhVKisBMQEIDZs2fjyZMnUtnjx48RERGBgIAAo3WOiIiIqLRKdBjr008/xSuvvIIaNWqgadOmAIDTp09DpVIhNjbWqB0kIiIiKo0ShZ3GjRvj8uXLWLduHS5dugQA6N+/PwYOHAgrKyujdpCIiIioNEoUdiIjI+Hi4oIRI0bolH/99de4c+cOpk6dapTOEREREZVWic7ZWbVqFXx8fPTKGzZsiJUrV5a6U0RERETGUqKwk5KSAjc3N71yJycnJCcnl7pTRERERMZSorDj4eGBQ4cO6ZUfOnQI7u7upe4UERERkbGU6JydESNGYMKECdBoNOjcuTMAYPfu3Xj//fd5B2UiIiKqUEoUdqZMmYJ79+5h7NixyM7OBgBYWlpi6tSpmD59ulE7SERERFQaJQo7CoUCCxcuxMyZM3Hx4kVYWVmhbt26UKlUxu4fERERUamUKOzksbW1RcuWLY3VFyIiIiKjK9EJykRERESVBcMOERERyRrDDhEREckaww4RERHJGsMOERERyRrDDhEREckaww4RERHJGsMOERERyRrDDhEREckaww4RERHJGsMOERERyRrDDhEREckaww4RERHJGsMOERERyZpJw87+/fvRo0cPuLu7Q6FQYPPmzTrThwwZAoVCofPzyiuv6NS5f/8+Bg4cCLVaDQcHBwwfPhwPHz4sx1EQERFRRWbSsPPo0SM0bdoUy5cvz7fOK6+8guTkZOnnhx9+0Jk+cOBAnD9/HnFxcdi2bRv279+PkSNHlnXXiYiIqJKoYsqFd+vWDd26dSuwjkqlgqurq8FpFy9exM6dO3H06FG0aNECALBs2TJ0794dS5Ysgbu7u9H7TERERJWLScNOUezbtw/Ozs6oWrUqOnfujHnz5qFatWoAgISEBDg4OEhBBwACAwNhZmaG33//Ha+99prBNrOyspCVlSW9zsjIAABoNBpoNJpS9zmvDWO0Vdlw7EUfu8pclGV3ypXKTOj8+yIpq7FXhs8QP+8cu6kVtQ8KIUSF+HZSKBTYtGkTevfuLZX9+OOPsLa2hpeXF65evYoPPvgAtra2SEhIgLm5ORYsWIBvvvkGiYmJOm05OzsjIiICY8aMMbis8PBwRERE6JVHR0fD2traqOMiIiKispGZmYkBAwYgPT0darU633oVes9Ov379pP83btwYTZo0gbe3N/bt24cuXbqUuN3p06dj0qRJ0uuMjAx4eHggKCiowDerqDQaDeLi4tC1a1colcpSt1eZcOxFH3uj8Jhy6FX5UJkJzG2hxcxjZsjSKkzdnXJVVmM/Fx5stLbKCj/vHLupx553ZKYwFTrsPK927dqoXr06rly5gi5dusDV1RWpqak6dXJycnD//v18z/MBnp4HpFKp9MqVSqVRV5yx26tMOPbCx56VK79QkKVVyHJcRWHssVemzw8/7xy7KftQFJXqPjv//PMP7t27Bzc3NwBAQEAA0tLScPz4canOnj17oNVq4e/vb6puEhERUQVi0j07Dx8+xJUrV6TXSUlJOHXqFBwdHeHo6IiIiAiEhobC1dUVV69exfvvv486deogOPjp7t0GDRrglVdewYgRI7By5UpoNBqMGzcO/fr145VYREREBMDEe3aOHTuG5s2bo3nz5gCASZMmoXnz5pg1axbMzc1x5swZ9OzZE/Xq1cPw4cPh5+eHAwcO6ByCWrduHXx8fNClSxd0794dbdu2xerVq001JCIiIqpgTLpnp2PHjijoYrCYmMJP4HR0dER0dLQxu0VEREQyUqnO2SEiIiIqLoYdIiIikjWGHSIiIpI1hh0iIiKSNYYdIiIikjWGHSIiIpI1hh0iIiKSNYYdIiIikjWGHSIiIpI1hh0iIiKSNYYdIiIikjWGHSIiIpI1hh0iIiKSNYYdIiIikjWGHSIiIpI1hh0iIiKSNYYdIiIikjWGHSIiIpI1hh0iIiKSNYYdIiIikjWGHSIiIpI1hh0iIiKSNYYdIiIikjWGHSIiIpI1hh0iIiKSNYYdIiIikjWGHSIiIpI1hh0iIiKSNYYdIiIikjWGHSIiIpI1hh0iIiKSNYYdIiIikjWGHSIiIpI1hh0iIiKSNYYdIiIikjWGHSIiIpI1hh0iIiKSNZOGnf3796NHjx5wd3eHQqHA5s2bdaYLITBr1iy4ubnBysoKgYGBuHz5sk6d+/fvY+DAgVCr1XBwcMDw4cPx8OHDchwFERERVWQmDTuPHj1C06ZNsXz5coPTFy1ahM8//xwrV67E77//DhsbGwQHB+PJkydSnYEDB+L8+fOIi4vDtm3bsH//fowcObK8hkBEREQVXBVTLrxbt27o1q2bwWlCCHz66aeYMWMGevXqBQD49ttv4eLigs2bN6Nfv364ePEidu7ciaNHj6JFixYAgGXLlqF79+5YsmQJ3N3dy20sREREVDFV2HN2kpKSkJKSgsDAQKnM3t4e/v7+SEhIAAAkJCTAwcFBCjoAEBgYCDMzM/z+++/l3mciIiKqeEy6Z6cgKSkpAAAXFxedchcXF2laSkoKnJ2ddaZXqVIFjo6OUh1DsrKykJWVJb3OyMgAAGg0Gmg0mlL3Pa8NY7RV2XDsRR+7ylyUZXfKlcpM6Pz7IimrsVeGzxA/7xy7qRW1DxU27JSlyMhIRERE6JXHxsbC2traaMuJi4szWluVDcdeuEWtyrgjJjC3hdbUXTAZY499+/btRm2vLPHz/mKqCGPPzMwsUr0KG3ZcXV0BALdv34abm5tUfvv2bTRr1kyqk5qaqjNfTk4O7t+/L81vyPTp0zFp0iTpdUZGBjw8PBAUFAS1Wl3qvms0GsTFxaFr165QKpWlbq8y4diLPvZG4THl0KvyoTITmNtCi5nHzJClVZi6O+WqrMZ+LjzYaG2VFX7eOXZTjz3vyExhKmzY8fLygqurK3bv3i2Fm4yMDPz+++8YM2YMACAgIABpaWk4fvw4/Pz8AAB79uyBVquFv79/vm2rVCqoVCq9cqVSadQVZ+z2KhOOvfCxZ+XKLxRkaRWyHFdRGHvslenzw887x27KPhSFScPOw4cPceXKFel1UlISTp06BUdHR9SsWRMTJkzAvHnzULduXXh5eWHmzJlwd3dH7969AQANGjTAK6+8ghEjRmDlypXQaDQYN24c+vXrxyuxiIiICICJw86xY8fQqVMn6XXeoaXBgwdj7dq1eP/99/Ho0SOMHDkSaWlpaNu2LXbu3AlLS0tpnnXr1mHcuHHo0qULzMzMEBoais8//7zcx0JEREQVk0nDTseOHSFE/lcwKBQKzJkzB3PmzMm3jqOjI6Kjo8uie0RERCQDFfY+O0RERETGwLBDREREssawQ0RERLLGsENERESyxrBDREREssawQ0RERLLGsENERESyxrBDREREssawQ0RERLLGsENERESyxrBDREREssawQ0RERLLGsENERESyxrBDREREslbF1B0gMoZa034zdRegMhdY1ApoFB6DrFyFqbtDRET/H/fsEBERkawx7BAREZGsMewQERGRrDHsEBERkawx7BAREZGsMewQERGRrDHsEBERkawx7BAREZGsMewQERGRrDHsEBERkawx7BAREZGsMewQERGRrDHsEBERkawx7BAREZGsMewQERGRrDHsEBERkawx7BAREZGsMewQERGRrDHsEBERkawx7BAREZGsMewQERGRrDHsEBERkawx7BAREZGsVTF1BwoSHh6OiIgInbL69evj0qVLAIAnT55g8uTJ+PHHH5GVlYXg4GB8+eWXcHFxMUV3iYiMpta030zdhUKpzAUWtQIahccgK1eBax+FmLpLRAZV+D07DRs2RHJysvRz8OBBadrEiROxdetWbNiwAfHx8bh16xZef/11E/aWiIiIKpoKvWcHAKpUqQJXV1e98vT0dKxZswbR0dHo3LkzACAqKgoNGjTAkSNH8PLLL5d3V4mIiKgCqvBh5/Lly3B3d4elpSUCAgIQGRmJmjVr4vjx49BoNAgMDJTq+vj4oGbNmkhISCgw7GRlZSErK0t6nZGRAQDQaDTQaDSl7nNeG8Zoq7Ix1dhV5qJcl2ewD2ZC598XCcfOsQMv1ncev+crxtiL2geFEKLCfkJ37NiBhw8fon79+khOTkZERARu3ryJc+fOYevWrRg6dKhOaAGAVq1aoVOnTli4cGG+7Ro6FwgAoqOjYW1tbfRxEBERkfFlZmZiwIABSE9Ph1qtzrdehQ47z0tLS4Onpyc++eQTWFlZlTjsGNqz4+Hhgbt37xb4ZhWVRqNBXFwcunbtCqVSWer2KhNTjb1ReEy5LSs/KjOBuS20mHnMDFlaham7U644do49S6vAufBgU3ep3PB7vmKMPSMjA9WrVy807FT4w1jPcnBwQL169XDlyhV07doV2dnZSEtLg4ODg1Tn9u3bBs/xeZZKpYJKpdIrVyqVRl1xxm6vMinvsWflVpxfMllaRYXqT3ni2F/ssb+I33f8njft2Iu6/Ap/NdazHj58iKtXr8LNzQ1+fn5QKpXYvXu3ND0xMRE3btxAQECACXtJREREFUmF3rPz3nvvoUePHvD09MStW7cwe/ZsmJubo3///rC3t8fw4cMxadIkODo6Qq1WY/z48QgICOCVWERERCSp0GHnn3/+Qf/+/XHv3j04OTmhbdu2OHLkCJycnAAAS5cuhZmZGUJDQ3VuKkhERESUp0KHnR9//LHA6ZaWlli+fDmWL19eTj0iIiKiyqZSnbNDREREVFwMO0RERCRrDDtEREQkaww7REREJGsMO0RERCRrDDtEREQkaww7REREJGsMO0RERCRrDDtEREQkaww7REREJGsMO0RERCRrDDtEREQkaww7REREJGsMO0RERCRrDDtEREQkaww7REREJGsMO0RERCRrDDtEREQka1VM3QGqeGpN+63E86rMBRa1AhqFxyArV2HEXhEREZUM9+wQERGRrDHsEBERkawx7BAREZGsMewQERGRrDHsEBERkawx7BAREZGsMewQERGRrDHsEBERkawx7BAREZGsMewQERGRrDHsEBERkawx7BAREZGsMewQERGRrDHsEBERkaxVMXUH5K7WtN9M3QUionJRGb/vrn0UYuouUDngnh0iIiKSNYYdIiIikjWGHSIiIpI12YSd5cuXo1atWrC0tIS/vz/++OMPU3eJiIiIKgBZhJ3169dj0qRJmD17Nk6cOIGmTZsiODgYqamppu4aERERmZgsws4nn3yCESNGYOjQofD19cXKlSthbW2Nr7/+2tRdIyIiIhOr9GEnOzsbx48fR2BgoFRmZmaGwMBAJCQkmLBnREREVBFU+vvs3L17F7m5uXBxcdEpd3FxwaVLlwzOk5WVhaysLOl1eno6AOD+/fvQaDSl7pNGo0FmZibu3buHKjmPSt1eZVJFK5CZqUUVjRlytQpTd6dccewcO8de+dy7d69E8z37Pa9UKo3cq4L5R+4u1+U9T2UmMKO5Fs0+/AVZRVzvv0/vUiZ9efDgAQBACFFgvUofdkoiMjISEREReuVeXl4m6I38DDB1B0yIY38xceyVV/WPTd2Dyqm4672s3+cHDx7A3t4+3+mVPuxUr14d5ubmuH37tk757du34erqanCe6dOnY9KkSdJrrVaL+/fvo1q1alAoSv/XSUZGBjw8PPD3339DrVaXur3KhGPn2Dn2FwfHzrGbeuxCCDx48ADu7u4F1qv0YcfCwgJ+fn7YvXs3evfuDeBpeNm9ezfGjRtncB6VSgWVSqVT5uDgYPS+qdVqk28IpsKxc+wvGo6dY3/RVJSxF7RHJ0+lDzsAMGnSJAwePBgtWrRAq1at8Omnn+LRo0cYOnSoqbtGREREJiaLsNO3b1/cuXMHs2bNQkpKCpo1a4adO3fqnbRMRERELx5ZhB0AGDduXL6HrcqbSqXC7Nmz9Q6VvQg4do79RcOxc+wvmso4doUo7HotIiIiokqs0t9UkIiIiKggDDtEREQkaww7REREJGsMO6W0b98+KBQKgz9Hjx7Nd76OHTvq1R89enQ59tw4atWqpTeOjz76qMB5njx5grCwMFSrVg22trYIDQ3VuylkRXft2jUMHz4cXl5esLKygre3N2bPno3s7OwC56vM63358uWoVasWLC0t4e/vjz/++KPA+hs2bICPjw8sLS3RuHFjbN++vZx6ajyRkZFo2bIl7Ozs4OzsjN69eyMxMbHAedauXau3ji0tLcupx8YTHh6uNw4fH58C55HDOgcMf68pFAqEhYUZrF+Z1/n+/fvRo0cPuLu7Q6FQYPPmzTrThRCYNWsW3NzcYGVlhcDAQFy+fLnQdov7fVHWGHZKqXXr1khOTtb5eeedd+Dl5YUWLVoUOO+IESN05lu0aFE59dq45syZozOO8ePHF1h/4sSJ2Lp1KzZs2ID4+HjcunULr7/+ejn11jguXboErVaLVatW4fz581i6dClWrlyJDz74oNB5K+N6X79+PSZNmoTZs2fjxIkTaNq0KYKDg5Gammqw/uHDh9G/f38MHz4cJ0+eRO/evdG7d2+cO3eunHteOvHx8QgLC8ORI0cQFxcHjUaDoKAgPHpU8DPv1Gq1zjq+fv16OfXYuBo2bKgzjoMHD+ZbVy7rHACOHj2qM+64uDgAwJtvvpnvPJV1nT969AhNmzbF8uXLDU5ftGgRPv/8c6xcuRK///47bGxsEBwcjCdPnuTbZnG/L8qFIKPKzs4WTk5OYs6cOQXW69Chg/jvf/9bPp0qQ56enmLp0qVFrp+WliaUSqXYsGGDVHbx4kUBQCQkJJRBD8vPokWLhJeXV4F1Kut6b9WqlQgLC5Ne5+bmCnd3dxEZGWmwfp8+fURISIhOmb+/vxg1alSZ9rOspaamCgAiPj4+3zpRUVHC3t6+/DpVRmbPni2aNm1a5PpyXedCCPHf//5XeHt7C61Wa3C6XNY5ALFp0ybptVarFa6urmLx4sVSWVpamlCpVOKHH37It53ifl+UB+7ZMbJff/0V9+7dK9Ldm9etW4fq1aujUaNGmD59OjIzM8uhh8b30UcfoVq1amjevDkWL16MnJycfOseP34cGo0GgYGBUpmPjw9q1qyJhISE8uhumUlPT4ejo2Oh9Srbes/Ozsbx48d11pmZmRkCAwPzXWcJCQk69QEgODhYFusYQKHr+eHDh/D09ISHhwd69eqF8+fPl0f3jO7y5ctwd3dH7dq1MXDgQNy4cSPfunJd59nZ2fj+++8xbNiwAp+dKJd1/qykpCSkpKTorFd7e3v4+/vnu15L8n1RHmRzU8GKYs2aNQgODkaNGjUKrDdgwAB4enrC3d0dZ86cwdSpU5GYmIhffvmlnHpqHO+++y7+85//wNHREYcPH8b06dORnJyMTz75xGD9lJQUWFhY6D2LzMXFBSkpKeXQ47Jx5coVLFu2DEuWLCmwXmVc73fv3kVubq7eHcldXFxw6dIlg/OkpKQYrF+Z17FWq8WECRPQpk0bNGrUKN969evXx9dff40mTZogPT0dS5YsQevWrXH+/PlCvxcqEn9/f6xduxb169dHcnIyIiIi0K5dO5w7dw52dnZ69eW4zgFg8+bNSEtLw5AhQ/KtI5d1/ry8dVec9VqS74tyYbJ9ShXc1KlTBYACfy5evKgzz99//y3MzMzExo0bi7283bt3CwDiypUrxhpCiZVk7HnWrFkjqlSpIp48eWJw+rp164SFhYVeecuWLcX7779v1HGUREnG/s8//whvb28xfPjwYi+vIq33/Ny8eVMAEIcPH9YpnzJlimjVqpXBeZRKpYiOjtYpW758uXB2di6zfpa10aNHC09PT/H3338Xa77s7Gzh7e0tZsyYUUY9Kx///vuvUKvV4n//+5/B6XJc50IIERQUJF599dVizVNZ1zmeO4x16NAhAUDcunVLp96bb74p+vTpY7CNknxflAfu2cnH5MmTC0zyAFC7dm2d11FRUahWrRp69uxZ7OX5+/sDeLqHwNvbu9jzG1NJxp7H398fOTk5uHbtGurXr6833dXVFdnZ2UhLS9PZu3P79m24urqWpttGUdyx37p1C506dULr1q2xevXqYi+vIq33/FSvXh3m5uZ6V8wVtM5cXV2LVb+iGzduHLZt24b9+/cX+y91pVKJ5s2b48qVK2XUu/Lh4OCAevXq5TsOua1zALh+/Tp27dpV7D2vclnneevu9u3bcHNzk8pv376NZs2aGZynJN8X5YFhJx9OTk5wcnIqcn0hBKKiojBo0CAolcpiL+/UqVMAoLNBmUpxx/6sU6dOwczMDM7Ozgan+/n5QalUYvfu3QgNDQUAJCYm4saNGwgICChxn42lOGO/efMmOnXqBD8/P0RFRcHMrPinwFWk9Z4fCwsL+Pn5Yffu3ejduzeAp4d0du/ene/z6AICArB7925MmDBBKouLi6sQ67g4hBAYP348Nm3ahH379sHLy6vYbeTm5uLs2bPo3r17GfSw/Dx8+BBXr17F22+/bXC6XNb5s6KiouDs7IyQkJBizSeXde7l5QVXV1fs3r1bCjcZGRn4/fffMWbMGIPzlOT7olyYbJ+SzOzatSvfwzv//POPqF+/vvj999+FEEJcuXJFzJkzRxw7dkwkJSWJLVu2iNq1a4v27duXd7dL5fDhw2Lp0qXi1KlT4urVq+L7778XTk5OYtCgQVKd58cuxNPDATVr1hR79uwRx44dEwEBASIgIMAUQyixf/75R9SpU0d06dJF/PPPPyI5OVn6ebaOXNb7jz/+KFQqlVi7dq24cOGCGDlypHBwcBApKSlCCCHefvttMW3aNKn+oUOHRJUqVcSSJUvExYsXxezZs4VSqRRnz5411RBKZMyYMcLe3l7s27dPZx1nZmZKdZ4fe0REhIiJiRFXr14Vx48fF/369ROWlpbi/PnzphhCiU2ePFns27dPJCUliUOHDonAwEBRvXp1kZqaKoSQ7zrPk5ubK2rWrCmmTp2qN01O6/zBgwfi5MmT4uTJkwKA+OSTT8TJkyfF9evXhRBCfPTRR8LBwUFs2bJFnDlzRvTq1Ut4eXmJx48fS2107txZLFu2THpd2PeFKTDsGEn//v1F69atDU5LSkoSAMTevXuFEELcuHFDtG/fXjg6OgqVSiXq1KkjpkyZItLT08uxx6V3/Phx4e/vL+zt7YWlpaVo0KCBWLBggc75Os+PXQghHj9+LMaOHSuqVq0qrK2txWuvvaYTEiqDqKiofM/pySO39b5s2TJRs2ZNYWFhIVq1aiWOHDkiTevQoYMYPHiwTv2ffvpJ1KtXT1hYWIiGDRuK3377rZx7XHr5reOoqCipzvNjnzBhgvQ+ubi4iO7du4sTJ06Uf+dLqW/fvsLNzU1YWFiIl156SfTt21fn3DK5rvM8MTExAoBITEzUmyandb53716D23je+LRarZg5c6ZwcXERKpVKdOnSRe898fT0FLNnz9YpK+j7whT41HMiIiKSNd5nh4iIiGSNYYeIiIhkjWGHiIiIZI1hh4iIiGSNYYeIiIhkjWGHiIiIZI1hh4iIiGSNYYeIiIhkjWGHiMpMx44ddZ6VVJhr165BoVBIzwwzVrv79u2DQqFAWlpavnUUCgU2b95c5DaJqPLgg0CJqMz88ssvxXowroeHB5KTk1G9enUAT0NKp06d8O+//8LBwaHE7RLRi41hh4jKjKOjY7Hqm5ubw9XV1ejtEtGLjYexiKjMPH+4qVatWliwYAGGDRsGOzs71KxZE6tXr5amP3sY69q1a+jUqRMAoGrVqlAoFBgyZIjBdr/77ju0aNECdnZ2cHV1xYABA5Camlqqvp89exadO3eGlZUVqlWrhpEjR+Lhw4fS9H379qFVq1awsbGBg4MD2rRpg+vXrwMATp8+jU6dOsHOzg5qtRp+fn44duxYqfpDRCXHsENE5erjjz9GixYtcPLkSYwdOxZjxoxBYmKiXj0PDw/8/PPPAIDExEQkJyfjs88+M9imRqPB3Llzcfr0aWzevBnXrl2TglFJPHr0CMHBwahatSqOHj2KDRs2YNeuXRg3bhwAICcnB71790aHDh1w5swZJCQkYOTIkVAoFACAgQMHokaNGjh69CiOHz+OadOm8bAbkQnxMBYRlavu3btj7NixAICpU6di6dKl2Lt3L+rXr69Tz9zcXDpc5ezsrHPOzvOGDRsm/b927dr4/PPP0bJlSzx8+BC2trbF7mN0dDSePHmCb7/9FjY2NgCAL774Aj169MDChQuhVCqRnp6OV199Fd7e3gCABg0aSPPfuHEDU6ZMgY+PDwCgbt26xe4DERkP9+wQUblq0qSJ9H+FQgFXV9dSH3I6fvw4evTogZo1a8LOzg4dOnQA8DR0lMTFixfRtGlTKegAQJs2baDVapGYmAhHR0cMGTIEwcHB6NGjBz777DMkJydLdSdNmoR33nkHgYGB+Oijj3D16tVSjY+ISodhh4jK1fOHcxQKBbRabYnbyzvkpFarsW7dOhw9ehSbNm0CAGRnZ5eqrwWJiopCQkICWrdujfXr16NevXo4cuQIACA8PBznz59HSEgI9uzZA19fX6lPRFT+GHaIqMKysLAAAOTm5uZb59KlS7h37x4++ugjtGvXDj4+PqXeU9SgQQOcPn0ajx49ksoOHToEMzMzncNtzZs3x/Tp03H48GE0atQI0dHR0rR69eph4sSJiI2Nxeuvv46oqKhS9YmISo5hh4gqLE9PTygUCmzbtg137tzRuRoqT82aNWFhYYFly5bhr7/+wq+//oq5c+eWarkDBw6EpaUlBg8ejHPnzmHv3r0YP3483n77bbi4uCApKQnTp09HQkICrl+/jtjYWFy+fBkNGjTA48ePMW7cOOzbtw/Xr1/HoUOHcPToUZ1zeoiofDHsEFGF9dJLLyEiIgLTpk2Di4uLdDXUs5ycnLB27Vps2LABvr6++Oijj7BkyZJSLdfa2hoxMTG4f/8+WrZsiTfeeANdunTBF198IU2/dOkSQkNDUa9ePYwcORJhYWEYNWoUzM3Nce/ePQwaNAj16tVDnz590K1bN0RERJSqT0RUcgohhDB1J4iIiIjKCvfsEBERkawx7BAREZGsMewQERGRrDHsEBERkawx7BAREZGsMewQERGRrDHsEBERkawx7BAREZGsMewQERGRrDHsEBERkawx7BAREZGsMewQERGRrP0//65YLjdr3csAAAAASUVORK5CYII=",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "initial_loss_histogram(losses_lorenz, \"Lorenz\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "72d0f5a0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "initial_loss_histogram(losses_bzh, \"Belousov-Zhabotinsky\", bins=30)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "d36047a5",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Kepler: -2.56e-05 ± 1.30e-03\n",
      "Lorenz: 2.90e-01 ± 2.07e+00\n",
      "Belousov-Zhabotinsky: 6.92e-02 ± 6.90e-03\n"
     ]
    }
   ],
   "source": [
    "def mean_pm_std(array) -> str:\n",
    "    return f\"{np.mean(array):.2e} ± {np.std(array):.2e}\"\n",
    "\n",
    "print(\"Kepler:\", mean_pm_std(losses_kepler))\n",
    "print(\"Lorenz:\", mean_pm_std(losses_lorenz))\n",
    "print(\"Belousov-Zhabotinsky:\", mean_pm_std(losses_bzh))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9db50dcf",
   "metadata": {},
   "source": [
    "### Kepler&Lorenz are fine, i.e. centered at zero. But Belousov-Zhabotinsky??"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "base",
   "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": "3.8.16"
  },
  "vscode": {
   "interpreter": {
    "hash": "d4d1e4263499bec80672ea0156c357c1ee493ec2b1c70f0acce89fc37c4a6abe"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}