old_remove_double_descent.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "from os.path import join as oj\n",
    "import sys, time\n",
    "sys.path.insert(1, oj(sys.path[0], '..'))  # insert parent path\n",
    "import seaborn as sns\n",
    "from sklearn.model_selection import train_test_split\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import pmlb\n",
    "from tqdm import tqdm\n",
    "from copy import deepcopy\n",
    "import pickle as pkl\n",
    "import pandas as pd\n",
    "import data\n",
    "import numpy.linalg as npl \n",
    "import numpy.random as npr\n",
    "from scipy.optimize import minimize\n",
    "# from process_results import process_results, aggregate_results, best_ridge\n",
    "from scipy.stats import ortho_group\n",
    "from sklearn.linear_model import LinearRegression, RidgeCV\n",
    "import scipy.linalg as scl\n",
    "import pickle\n",
    "%matplotlib inline\n",
    "%load_ext autoreload\n",
    "%autoreload 2\n",
    "\n",
    "save_dir = 'results'\n",
    "os.makedirs(save_dir, exist_ok=True)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "def loc_calc_thetahat(l):\n",
    "    inv = npl.pinv(X_train.T @ X_train + l * np.eye(d))\n",
    "#     inv = eigenvecs @ npl.pinv(np.eye(d) + np.diag(eigenvals)) @ eigenvecs.T\n",
    "    return inv @ X_train.T @ y_train\n",
    "\n",
    "def loc_mdl1_loss(l):\n",
    "    thetahat = loc_calc_thetahat(l)\n",
    "    mse_norm = npl.norm(y_train - X_train @ thetahat)**2 / (2 * var)\n",
    "    theta_norm = npl.norm(thetahat)**2 / (2 * var)\n",
    "    eigensum = 0.5 * np.sum(np.log((eigenvals + l) / l))\n",
    "    return (mse_norm + theta_norm + eigensum) / n\n",
    "\n",
    "\n",
    "def loc_mdl1_loss_ret_all(l):\n",
    "    thetahat = loc_calc_thetahat(l)\n",
    "    mse_norm = npl.norm(y_train - X_train @ thetahat)**2 / (2 * var)\n",
    "    theta_norm = npl.norm(thetahat)**2 / (2 * var)\n",
    "    eigensum = 0.5 * np.sum(np.log((eigenvals + l) / l))\n",
    "    return mse_norm/n , theta_norm/n,  eigensum / n\n",
    "\n",
    "def g(a, b):\n",
    "    val = np.sqrt(a*(1+np.sqrt(b))**2+1)\n",
    "    val -= np.sqrt(a*(1-np.sqrt(b))**2+1)\n",
    "    val = val**2\n",
    "    return(val/4)\n",
    "\n",
    "def thm3(gammas, snr):\n",
    "    delta = g(snr, gammas)\n",
    "    bnd = gammas * np.log(1+snr-delta)\n",
    "    bnd += np.log(1+gammas * snr - delta)\n",
    "    bnd -= delta/snr\n",
    "    return(bnd)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "cs=sns.color_palette()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Load Results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "file_name = 'mdl_results_d_100_200.txt'\n",
    "file = open(file_name, 'rb')\n",
    "results = pickle.load(file)\n",
    "# results"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Plot results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib as mpl\n",
    "label_size = 20\n",
    "mpl.rcParams['xtick.labelsize'] = label_size \n",
    "mpl.rcParams['ytick.labelsize'] = label_size \n",
    "mpl.rcParams['axes.labelsize'] = label_size\n",
    "mpl.rcParams['axes.titlesize'] = 14\n",
    "mpl.rcParams['figure.titlesize'] = 10\n",
    "# mpl.rcParams['lines.markersize'] = 8\n",
    "mpl.rcParams['grid.linewidth'] = 2.5\n",
    "mpl.rcParams['legend.fontsize'] = 20\n",
    "# pylab.rcParams['xtick.major.pad']=5\n",
    "# pylab.rcParams['ytick.major.pad']=5"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAjgAAAGoCAYAAABL+58oAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4xLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvDW2N/gAAIABJREFUeJzs3Xd4lFXexvHvSackARJ6C4SOUkMRpa+CFVBWRGVFVlGx97a+iruWdVdUFFSsa0NFXViKiIoi0quAECKhhNBDCARC+nn/mCRkMpNAIJmZJPfnunKFnPM8k1+4RnJ7nlOMtRYRERGRysTP2wWIiIiIlDUFHBEREal0FHBERESk0lHAERERkUpHAUdEREQqHQUcERERqXQUcERERKTSUcARqeSMMVHGGFvkY78xJriY69e7uT6qhNfKNsacMMbsMcasMMZMM8YMMsaYYl7/mSL3Dyjjn7e1MeYlY8xyY8whY0yWMSbVGLPRGDPZGNPXXW3GmOrGmDuMMfPyfpZ0Y8xxY0y8MWa6MWZYCT/Tz27+XoYXc+39bq595gxezxpjcowxKcaYlcaYp40xtcvkL02kElLAEama6gM3FG00xvwJ6FzK1/IHqgONgJ7ArcCPwNL8YOQJxpggY8zLwFbgYaAXEAkEADWB84C7gV+A5kXu7QvEAVOBS3H8LMFADaAlcB0wE/jVGNPkDEt6wE2N/sC9pf3ZCvEDwoEewDPAZmNMh3N4PZFKK8DbBYiI19wHvF+k7cFSvsZq4AugGo4gcAWOUAHQG1hujLnAWrvjXAo9HWOMH/ApMLJQcwYwC9iEIxi0AoYAdYvc2wP4Lu9nyLcA+DWv7Uoc4QigD/CjMaa3tfbIacrqa4zpbq1dU6htJEXC1Rl6OO9zGDAM6JT3dQPgIyDmLF5TpFJTwBGpenJwjLqcb4y52Fr7PUDeSMCQIteczu/W2n/nf2GMqQ68DdyY11Qf+BDoXzalF2sszuFmG3CptXZb4YuMMUF516blfW2Ad3EON7daa98tdM/fgA+Av+Q1tQGexTEa5E7hv7sHcB4pe8DNNadV5O/4eWAz0CKvqbsxpkV5h0iRikaPqESqnlmF/vxAkT/nzzGZeTYvbK1NA24GNhZq7meM6Xk2r1cKRUeeRhcNNwDW2kxr7TRr7cG8pos4NRoCsLxwuMm7Jzfv9Y8Xah5X3BwmIBHIH7X5szGmMYAx5iIcj/DgLP9+8+pJL/T6+Rqe7euJVFYKOCJVz0bg+7w/DzHGtDfG1OPUqMtWYO7Zvri1NhvXR1+Dz/b1TscY0xAoPA/lN2vt6jO8vejI0ufuLrLWJuGYV5SvOiU/FpqU9zmQUyM9D7rpLzVjTAjQvUjzvrN9PZHKSo+oRKqml4GLcYzY3I/jF2T+iMSrgD3H199a5OsznZh7Noq+dmwp7m1U5OudJVxbtK/ovYV9CbwINAXGG2OmA1fl9a2w1i4tZkGWW8aYh/L+mD8Hp0Wh7jV6PCXiSgFHpAqy1n5njPkd6AiMAU7kdR0G/gOMOsdvUfS391kHJmNMHxyTe4vaZK2df7avW56stdnGmNeBl4DawBxOjZifzejNv4ppP4hjTpGIFKGAI1J1TQLeA0LyPgDestaeLM3oQjHaFvk68Rxe6xLgaTft/wHmu3ntdqV47b1Fvo4q4dqifUXvLWoa8BQQyqlRpl3A12dYmzsWSMWxpH0eMNlae/gcXk+k0tIcHJGq61PgQKGvM4E3zvVFjTEBOCYaF/aju2vLgrV2H45VRfk6G2OKzlEpzqIiX7sduTLGROA8jygNxxL5kuo6iutcpMnW2pwzrK3wa5m8Dz9rbbi1toe19mmFG5HiKeCIVFHW2gxgSqGm6dba/efymsaYajh+qZ9fqPlna+2qs31Na+0zhX7BF/4YW+iyoo99PjPGtCjShjEm0BgzPm9SNTj2uim84usCY8zNRe7xA/6NY7PAfO/n/f2dzqs4loQDHMOxJF1EPECPqESqtsmc+gW/8izu75g3ATYEiMZ5oz9wjBAVHc0p6m1jTKqb9r3W2qvctLvzAXA5MCLv6zY4dvktvNFfa05t9LcAwFprjTG3AD9zai+c940x1wJLcUy8vhLnpeR/4P6RmQtr7U5jzMU4dh/eb609doY/j4icIwUckSos7zHKWe/JgmOpdHHLpZcAN1prd57mNdoU0x5ZTLsLa22uMWY0jsm4d+GY5ByC45FTiROmrbUrjTFDgemcWhk1NO+jqGXAKGttcilq++lMrxWRsqOAIyLnKhfHsQgpOCb8rsfxuMujv9jzHhndY4yZAtwC9MMxqhSeV98OHCM1X+KY7Fv43l+MMW2Am3CM2HQB6gDZOFYqrcARgGbnbfwnIj7OWHuu212IiIiI+BZNMhYREZFKRwFHREREKh0FHBEREal0FHBERESk0qmSq6giIyNtVFSUt8uoENLT0wkJCTn9hSI+RO9bqWj0nj1za9asSbLW1j3ddVUy4ERFRbF6dYm7rEue2NhY2rUrzdE+It6n961UNHrPnjljzK7TX6VHVCIiIlIJKeCIiIhIpaOAIyIiIpWOAo6IiIhUOgo4IiIiUuko4IiIiEilo4AjIiIilY4CjoiIiFQ6CjgiIiJS6SjgiIiISKWjgCMiIiKVjgKOiIiIlJkjJzKZt3EfT83cRHpWjtfqqJKHbYqIiEjZOJ6RzaodySyNT2LJtsNs2X8Max19l57XgD6tIr1SlwKOiIiInLH0rBzWJaSwND6JpfGH+W13Ctm51u21S+MPK+CIiIiI78nOyWXT3mMs2ZbEsvjDrNqZTEZ27hnduyQ+iYdoW84VuqeAIyIiIgWstcQdOM6SbY4RmhXbD5OakV2q12jfMIwLoyO4qLV3Rm9AAUdERKRKs9ayO/kkS/IeOS2LTyLpeGapXqNFZA36REfQJzqSC6IjqFMjqJyqPXNVKuAYY64ErmzVqpW3S6k0oqKiANi5c6dX6xARkTN38Fg6S+MPF4zS7Ek5War7G4SF0KeVI9D0iY6gUa1q5VTp2atSAcdaOxuYHRMTc6u3a5HiJSYm8p///If169ezbt06tm/fjrWWP/74g5LC6cmTJ3nxxRf5/PPP2bVrF2FhYQwYMICJEyfSvn17t/ckJyfz7LPPMnPmTPbt20dERARDhw7l2WefpUmTJuX1I4qIeFRKWibLtycXTAzedvB4qe6vXT2QC6JPBZoWkTUwxpRTtWWjSgUcqRhWr17N3/72N4wxtGjRgvDwcFJSUkq8JyMjg4svvpglS5YQExPDvffey+7du5kxYwZz585l4cKF9OrVy+mew4cP06dPH+Li4hg0aBDXXXcdsbGxfPDBB8ydO5dly5bRsmXL8vxRRUTKRVpmNqt2HmFp3gjNpr1HC5Zun4kaQf70bFGHC1s5Hjm1bxCGn59vB5qiFHDE58TExPDLL7/QuXPnglGYRYsWlXjPpEmTWLJkCSNHjuSLL77Az8+xh+WoUaMYPnw448aNY+PGjQXtAE888QRxcXE88MADvPzyywXtkydP5t5772XChAnMnz+/fH5IEZEylJmdy/rdKQUrndbtPkJWzpknmiB/P7o1r8WF0ZH0aRVBpya1CPSv2HsBK+DIaVlrmTJlCm+++Sbx8fFEREQwYsQInnvuuXL5fk2aNCnV4yFrLW+99RYAL730klOIGTZsGH379mXx4sUsWrSIgQMHAnD8+HE+/vhjatSowTPPPOP0enfddReTJk3iu+++Y/v27RrFERGfk5Nr+X3v0YJ5NKt3HuFkKXYN9jPQqUkt+kRHcGGrSLo3r01IoH85Vux5CjhnIeqxud4uoVR2vnj5Od1/3333MXnyZBo2bMj48eMJDAxk1qxZrFixgszMTIKCvDtbPj4+noSEBNq0aUOLFi1c+i+99FIWL17MwoULCwLO8uXLOXnyJJdccgmhoaFO1/v5+TFkyBCmTZvGTz/9pIAjIl5nrWXbweMFgWb59sMcSy/d0u12DUK5IDqCC6Mj6dmyDmEhgeVUrW9QwJESrVu3jsmTJxMdHc3KlSupU6cOAM899xwDBw5k3759NG/evOD6lJQUXn311VJ9j+HDh9OlS5ezrnHr1q0AtGnTxm1/69atAYiLizune0REPGl3chrL4g8XLN8+lJpRqvubR1QvWLrdu2UEdUODy6lS36SAIyX65ptvAHjyyScLwg1ASEgIL7zwQsGISL6UlBQmTpxYqu8RFRV1TgHn6NGjAISHh7vtz28vPFH5bO4RESlPh1IzWBqfVBBqdieXbul2vdDggknBfaIjaFK7ejlVWjEo4EiJNm/eDED//v1d+i666CL8/Z2f2UZFRWFLM1VfRKQKiz90nC9X7Wb+ht3sSokv1b3h1QK5oGVEwX400XV9f+m2JyngnIVzndNSkRw/7tgroX79+i59AQEBREZ6bxvufPmjLfmjMkXlt9eqVeuc7hERKQvpWTnM37Sf6SsTWLEj+YzvqxboWLqdPzG4fcMw/H1t6XZ2JuzfAPU6QJB3R5AUcKRENWvWBODAgQMuk22zs7NJSkpyWvHkjTk4bds6DnIrbr7MH3/8ATjPtzmbe0REzsW2g6lMX7mbr9cmkpKWddrrA/0NXZvVLli63blJLYICfGzpdup+2L0SElc6Pu9dDzkZMGYmRA88/f3lSAFHStShQwc2b97MokWLXALOr7/+Sk6O87JEb8zBiY6OplmzZsTFxbFjxw6XlVTffvstAIMGDSpo6927N9WqVWPJkiWkpqY6raTKzc1lwYIFAC5zjERESiM9K4d5G/fx+crdrNxZ8miNMXB+4/CC3YJ7RNWhWpAPLd3OyYL9GyFxlSPM7F4JRxPcX5u4SgFHfNuIESP46quveO655xg2bFjBROP09HQef/xxl+u9MQfHGMPtt9/OE088wSOPPOK00d+sWbNYvHgxHTp0cJpHVLNmTcaMGcO0adN45plnnDb6e+ONN9i5cydDhgzREnEROStxB1L5bEUC36xNPO1y7h5RtenfJJAxg7oQXt2Hlm5nnoDtP+eN0KyCPWsh+wwnPu9eWa6lnQkFHClRt27duPvuu3n99dc577zzGDlyZME+OLVr16Zhw4bl8n3Hjh1b8OfY2FgAHn300YKRlltuuYWLLrqo4JoHHniAOXPm8NVXX9GrVy8GDx5MQkICM2bMoHr16rz//vtOGwACPP/88/z8889MmjSJ9evX07NnT7Zs2cKsWbOoV68eU6ZMKZefTUQqp5OZOczduI/pKxNYs+tIideGVwvkmm5NGN2zKa3rhxIbG+tb4QbgxCH4/PrS3xfaCMIalX09pWSq4oqXmJgYu3r1am+XUSHExsbStm1bpkyZwpQpU9i+fXvBTsbPP/88nTt3Bsr+NPHTrQT44IMPnEIQQFpaGi+++CLTp08nISHB6bDNDh06uH2d5ORkJk6c6HTY5qWXXqrDNiu42NhY2rVr5+0ypIqI3X+M6SsS+GbdHlJPM1rTs0Udru/ZjKHnNXDaOdjj79kThx2jMokrIbwpxNzseo218O82cOJg8a/jFwgNO0PTntCkh+NzePn+22mMWWOtjTntdQo4UhL9opCKSO9bKW9pmdnM2eAYrVmXUPJ+WbWrO0ZrruvZjFb1arq9plzfs7k5cHAL7F5xav5McqEl6U17w1+/c3/v5zdA7JxTX9dsAE17QJOejjDTsAsEhpRP3cU404CjR1QiIiJnaPPeY0xfmcDMdXtIzSh5tKZ3yzqM7tmMIR0bePacp7RkSFx9amXTnrWQmVr89XvXOZZ3B7g5dqfDMAhr7AgzTXs6RnsqyF47CjgiIiIlOJGRzZwNe/ls5W5+213yaE2dGkH8uXsTRvVoSsu67kdrys23j0L8Qkgq5REzORmO1VFNurv2dbrW8VEBKeCIiIi4sWnPUaavTGDW+r0cP81oTZ/oCK7v1YyLO9QnOKAcR2uy0ot/JHRwy5mHG+MP9TtC016OkZk6rgcVV3QKOCIiInmOZ2Qz+7e9TF+ZwIZE9zud54usGcTI7k25rkdToiJrlH0xublweFveo6YVsHsVHD8AD8eDn5sN/5r2hB2L3L9WtTqnHjM16QmNu0FQOdTsQxRwRESkytuYeJTPVibwv/V7OJGZU+K1fVtHMrpnM/7Uvn7Z7iyckQp71pzaRC9xFaS7eSR2eBvUdbPLepOejs/Gz3FUQn6YadoT6rSsMHNnyooCjoiIVEnHM7KZtX4P01cmsGnPsRKvjawZzLUxjrk1zSPKcOQjLRmWTiZq02w4uh1s7unvSVzpPuA06w1/mQWNu0NwqGt/FaOAIyIiVYa1lg2Jjrk1//ttL2kljNYYA31b1+X6nk0Z3L4+gf5lOFqTmwvrP4Hvn4aTyZz5QmsDR3a57woJg5YDyqS8ykABR0REKr1j6VnMWr+X6SsS2Lyv5NGauqHBjIppyqgeTWlap5xOxP72EVj1zumvCw6HJjGnNtJrEgMh4eVTUyWjgCMiIpWStZb1u1OYvjKB2b/t42RWyaM1/dvUZXTPZgxqV69sR2vc6X4TrH7P9ZFUZFvnjfQi27qfUCynpYAjIiKVytGTWcxav4fPViQQu7+EDe6A+mGO0ZprezSlSe1yGq1xp8H50PM2WPEmBFTjUPux1L30Eahex3M1VHIKOCIiUuFZa1mb4BitmbNhL+lZxU/W9TMwoG09RvdsxsC2dQkoz9GaE0lQI9J938AnHCun+j/C4QMnqatwU6YUcEREpMI6mpbFf9clMn3lbrYeKHm0pmF4CNfmjdY0rlWtfAvLOA6LXoQV0+CWH6BhJ9drQsJg+BTHnw/Elm89VZACjoiIVCjWWtbsOsJnKxOYu2EfGdklj9YMaucYrenfppxHaxzFweaZMP8JSN3raJv7IIz7TnNpPEwBR85JVFQUADt37vRqHSJS+WVm5zJz3R7e/XU7cQeOl3hto/AQRvVoxrU9mtAwvJxHa/IlbYN5D8H2n5zbE1fC+k+h2xjP1CEAKE6Kz1myZAmPPPIIPXr0oG7dugQHB9OiRQtuueUWtm3bVux9J0+e5Omnn6Zt27aEhIRQr149rr32WrZs2VLsPcnJydx3331ERUURHBxMo0aNGDduHImJieXxo4nIWUjLzOb9X3fQ/18/8cjXG4oNN/5+hos71OeDsT1Y/Ogg7v1Ta8+Em8w0+PHv8OYFruEGIKwJ1Khb/nWIE43giM+55pprOHToEH369OGGG24gICCAZcuW8d577/H555/z/fffc8EFFzjdk5GRwcUXX8ySJUuIiYnh3nvvZffu3cyYMYO5c+eycOFCevXq5XTP4cOH6dOnD3FxcQwaNIjrrruO2NhYPvjgA+bOncuyZcto2bKlJ390ESnkaFoWHy3byQdLd5J8IrPY6xrXqsbonk35c0xT6oed+ZZ5ZSJ2Hsx/FFISXPv8AuCCu6D/I5X+3CdfpIAjPuf+++9nzJgxNGrUyKn9+eef58knn2T8+PFs3LjRqW/SpEksWbKEkSNH8sUXX+CX96x71KhRDB8+nHHjxrFx48aCdoAnnniCuLg4HnjgAV5++eWC9smTJ3PvvfcyYcIE5s+fX44/qYi4czA1nfd+3cGnyxOKPcXb389wcfv6jO7VjItaReLv5+Fzlo7shG8fhbhi/o2I6guXvwx123q0LDnFWGu9XYPHxcTE2NWrV3u7jAohNjaWtm3bMmXKFN58803i4+OJiIhgxIgRPPfcc3Tu3BnwzBycnJwcQkNDOXnyJElJSURERACOCYdRUVEkJCSwfft2WrRo4XRfv379WLx4MQsXLmTgwIEAHD9+nHr16uHn58e+ffsIDT11bktubi4tW7Zk165dxMfHaxSnAoqNjaVdu3beLkNKaXdyGm//Es+XqxPJLGbicHCAH6N7NuOWvi08u29NvuwMWDIZFv8bstNd+2s2gCHPwXnXlOpwS71nz5wxZo21NuZ012kE51w8c5bbZTfsDLf94r7v7X6w77ezrOfo2d13Gvfddx+TJ0+mYcOGjB8/nsDAQGbNmsWKFSvIzMwkKCioXL5vUcYYAgIcb1l/f/+C9vj4eBISEmjTpo1LuAG49NJLXQLO8uXLOXnyJJdccolTuAHw8/NjyJAhTJs2jZ9++kkBR6ScxR1I5c2f4/nfb3vJyXX/P92hwQGMuaA54y5qQWTNYA9XWMjn18O2H1zbjT/0ug0GPO5Y/i1ep4AjJVq3bh2TJ08mOjqalStXUqeOYyOq5557joEDB7Jv3z6aN29ecH1KSgqvvvpqqb7H8OHD6dKly2mvmzFjBqmpqfTu3ZtatWoVtG/duhWANm3cnK4LtG7dGoC4uLhzukdEytbahCNM/SmeH7YcKPaayJpBjLuoBTf2bk5YSKAHqytGz9tcA07T3nD5vx27E4vPqNABxxgzAXgYaAj8DtxnrV3s3aoql2+++QaAJ598siDcAISEhPDCCy8UjIjkS0lJYeLEiaX6HlFRUacNODt27ODuu+8mICCASZMmOfUdPeoYuQoPdz+ilt+ekpJyTveIyLmz1rJk22Gm/LSNZdsPF3td41rVuK1/S66NaUpIoH+x13lcm0ug3RUQOweqR8LFz0Ln0drjxgdV2IBjjBkFvAZMAH7N+/ytMaaDtdbNdHY5G5s3bwagf//+Ln0XXXSR06MicISVsp7XdfDgQS699FIOHTrElClTXFZQiYjvy821LNh8gKk/b2NDYvGP01vVq8kd/aO5qkuj8j/wsiSp+yG0gfu+oS9AWCPHUQvVanu2LjljFTbgAA8AH1pr88+bv9sYMxS4A3jcIxWUx5yX4ubmeMnx4479JurXr+/SFxAQQGRkMWeslJGDBw8yaNAgtm7dymuvvcaECRNcrskfbckflSkqv73wY62zuUdESi8rJ5dZ6/fy1qJ4th0sfnO+zk3CuWNAKy7pUB8/T6+IKix1P3z3JMR9B3etgrCGrtfUagaX/cvztUmplFvAMcaMBPoDXYDOQCjwqbX2xhLuaQI8CwwFIoB9wExgorX2SKHrgoDuwL+LvMQCoE8Z/hhVXs2aNQE4cOCAy2Tb7OxskpKSaNKkSUFbWc7B2bdvH4MHDyY2NpYpU6a4DTcAbds6lmEWN1/mjz/+AJzn25zNPSJy5tKzcvhi1W6m/bKdPSkni72uT3QEEwa04sJWEZhSrDoqcznZsHIa/PQ8ZOadabXgSRj5vvdqknNSniM4f8MRbI4DiUCJ69+MMdHAUqAeMAuIBXoC9wJDjTEXWmvzH9hGAv5A0ZlpB4A/ldUPINChQwc2b97MokWLXALOr7/+Sk5OjlNbWc3BSUxMZNCgQWzbto233nqL8ePHF3t/dHQ0zZo1Iy4ujh07drispPr2228BGDRoUEFb7969qVatGkuWLCE1NdVlmfiCBQsAXOYYiUjJjqVn8fGyXXywZAdJx4vfnO/iDvWZMCCars184BFPwnLHeVEHNjm3b/oauv0FWg7wRlVyjsrzAef9QBsgDMdjo9OZiiPc3GOtHW6tfcxaOwh4BWgLPFdulUqxRowYAThWTSUnJxe0p6en8/jjrk8C8+fglOZj7NixTq+xa9cu+vXrR3x8PO+//36J4QYcy8dvv/12AB555BFyc0/tnzFr1iwWL15Mhw4dnOYR1axZkzFjxnDixAmeeeYZp9d744032LlzJ0OGDNEScZEzlHQ8g5fmx3LhCwv513db3YYbfz/DiK6N+e6+frzzlxjvh5sTSTDzTnh/iGu4Aah/PgSHurZLheCRjf6MMQOAnyjmEVXe6M02YCcQba3NLdQXiuNRlQHqWWtP5D2iSgNGW2tnFLp2CnCetdZ1Rmwh2ujvzMXGxjJ16lRef/11GjZsyMiRIwv2walduzZ79uwhKCioTDf6a9GiBTt37qR79+5cccUVbq8ZO3ZswUGf4DiqYdCgQSxdupSYmBgGDx5MQkICM2bMICgo6IyOaujZsydbtmxh1qxZ1KtXj6VLlxIdHV1mP5d4jjZN85zEI2m888t2Pl+1u9hTvYMC/Lg2pgm39YumaR0vbM5XVG4OrPkQfnwW0t2slAwOg4FPQo9bwN8zU1X1nj1zFW2jv/znAAsKhxsAa22qMWYJcAnQG/jRWptpjFkDXAzMKHT5xcDXnii4Knnttddo06YNU6ZM4e233y7Yyfj5558v2Mm4LOWHpTVr1rBmzRq31wwYMMAp4AQHB/P999/z4osvMn36dF555RXCwsIYPnw4EydOpEOHDi6vERERwbJly5g4cSIzZ85k8eLFREREcPPNN/Pss886zS0SEWfbDqby5s/bmbV+D9nFbM5XMziAG3s3Z9xFUdQL9fAZUcXZs8bxOGrvOvf9nUbBxX+HUNeFFVKx+MoIzr+Ah4CHrLUvu+l/A7gTmGCtfTOvbRTwMY7l4UuA24G/Ah2ttbvcvMZ4YDxAo0aNuv/4449l88NVcklJSeW+UkqkrOl9W37iktL5YmMKS3edoLjfHuHBfgzrUIsr24URGuwbe9j4ZRyl7oY3qRU/E+Om8oywFuyPeYST9bp5oTq9Z0ujffv2FWoEJ3+3teLWXee3F6zZtdZ+YYyJwDGZuSGwCbjMXbjJu34aMA0cj6g0FHhmNGwqFZHet2XLWsuy7YeZ+lM8v25LKva6huEhjO/XklE9mlI9yFd+veB4JDWlJxze5toXWAMGPEZw7zto7u+9nZL1ni17PvQOLD1r7VQck5NFRKSM5eZafthygKk/x7N+d/G7ereMrMHtA6IZ3qUxQQE+uKOvnz/0uh3mPeTc3mE4DHkewht7py4pV74ScPJHaIo7vTK/Xfvmi4iUs+ycXOZs2MfUn7cRd6D4zfk6NgrjzoGtGNKxAf7e3JzvTMSMg3WfwL71UCfasVFfq8HerkrKka8EnK15n4vbVa113medfCgiUk7Ss3KYsSaRab/Eszu5+M35erWow4SBrejXOtK7m/MVZS0c2+t+RMbPHy6fBNsXQp97IMCLJ5KLR/hKwPkp7/Mlxhg/N8vEL8SxLHy5N4oTEanMUtOz+HRFAu/9uoNDqRnFXje4XT0mDIyme/M6xV7jNQdjHY+gknfAXSshqIbrNU26Oz6kSvCJgGOtjTfXsnEXAAAgAElEQVTGLMCxFPxO4PVC3ROBGsDb1toT5/J9jDFXAle2atXqXF5GRKRSOHw8gw+X7uQ/S3dyLD3b7TV+Bq7o1Ig7BkTTvmGYhys8AxnH4ZeXYNkUyM37GRa9BBeXbkd1qXzK8yyq4cDwvC/zj2S9wBjzYd6fk6y1hWd8TcBxVMNkY8xgYAvQC8ceOXHAk+dak7V2NjA7Jibm1nN9LRGRimpvykneWbyd6SsTSM8qZnM+fz+u6d6E2/q1JCrSzWiIt1kLW/4H8x+HY3uc+5a9AV2uh7ptvVOb+ITyHMHpAtxUpK1l3gfALhx73wAFozgxnDps8zIcOxi/RpHDNkVEpPS2HzrOW4vi+e+6PWTluN/FpnqQPzf0asYtfVtSP8xHNucr6nA8zHsY4ovZz6xFP/AP8mxN4nPKLeBYa58BninlPbuBm8ujHhGRqmrTnqNM/Xkb327aT3F7u9aqHsjYPlGM7RNFreo+Gg72b4KVb8Nvn0OOm4M8wxrD0Beg/VXgS5OfxSt8Yg6OiIiUrb0pJ5m3cR9zNuwrcQ+b+mHB3Nq3JaN7NqNGsA/+SsjNga3fwoq3YOdi99f4BcAFd0K/RyC4pmfrE5/lg+9mERE5GweOpReEmjW7Sn6qHxVRndv7RzOiW2OCA3zjOAUXh7bCpyMhJaH4a6L6wmX/hnraBVicKeCIiFQAi/84xP/N+p0Dx9KpHhRAjWB/auR/Dg4gNT2btQlHin0Ela9dg1DuHNiKy85v6Pub89WOgqx0933hzWDwU3D+n/U4StxSwBER8XHbDx1n/EdrOJmVA0BaZg5JxW8w7FavFnW4vX80A9rW9a3N+UoSEOzYgXjRi6faml8EvW6DtpeBv36FSfGq1LtD++CISEWTmZ3LfV+sLwg3pdGlaS2u6NSQy85vSKNa1cqhunOUfgx+mw4bvoSbZkNQdddrYsbB8qnQ4SroeRs07OT5OqVCqlIBR/vgiEhF88oPcWxIPHr6C/N0ahLO5ec7Qk3TOm4Cgy84HA8rp8G6TyEz1dG28UvoPtb12tD68OBW9+FHpARVKuCIiFQkS+OTeGtRvFPb0I4NeHZYR45nZJOWmcOJjGxOZGaTmW3p0DCMZhE+GgSshe0/wfK34I8FQJHJQsvfgm43uZ9Po3AjZ0EBR0TEB2Xn5PLo1xucJg03CAvhxWvOp1b1IOp5r7TSyTzh2LdmxduQtLX46w5vc6ya0mooKSMKOCIiPmhJ/GGnE72NgUnXdvbdTfiKOrILVr0Daz+C9BIesdWo55hnEzPO8ThKpIwo4IiI+KBZ65zPVxrRpTF9WkV6qZpS2LXMcRbU1nlg3Z9zBUDDLtD7Dug4wrFaSqSMKeCIiHhYelYOP2w5wK7DacVe893v+52+vqZ7k/Iuq2xsngmxc9z3GX/oMMwRbJr00P41Uq4UcEREPMRay9yN+3hhXix7Uk6e/oY89UKD6d0yohwrK0M9xzvm2xSeRFytDsTcDDF/hfDGXitNqpYqFXC0D46IeMvGxKM8O+d3Vu0s+QgFd67q3Mh3dh22FnavAOMHTXu69kdEQ+tL4I/voP550Ot2OH8kBPrgPjxSqVWpgKN9cETE0w4eS+el77by9drE0x6j4E5oSAA39Ykq87pKLTsDNn3jOPRy33po1gfGfev+2oFPwIX3QPML9RhKvKZKBRwREU9Jz8rhvV93MOWnbaRluu5CHOBnGNG1MZGhpybYFg1ANYP9GXpeA+9u2Jd6AFa/B6vfhxOHTrUnLIV9v0HDzq73NOriufpEiqGAIyJShqy1fLtpP8/P20LiEffzbAa1q8cTl7WnVb2aHq6uFPascWy+9/t/ITfL/TUr3obhUz1bl8gZUsARESkjm/Yc5dk5m1m5I9ltf6t6NXnqig70b1PXw5WdoZws2DzLEVwSV5ZwoYE2Q6DTtR4rTaS0FHBERM5RelYOL83fygdLd7idZ1OreiAPXNyG63s2I8Dfz/MFns6JJFjzAax6D1L3FX9dUCh0vRF63uqYTCziwxRwRETOQdyBVO6Zvo7Y/akufQF+hjEXNOfewa19dwfilN3wenfIySj+mjotHauhOo+GkDDP1SZyDhRwRETOgrWWj5fv4rm5W8jIdt2xd2Dbujx5eQffnmcDUKupY6Kwu0dS0YOg1x3Q6k/g54MjTyIlUMARESmlpOMZPPrVBn6MPejS17hWNf4x4jwGtvWx4zAz04o/lbv37fBVXsAJrO4Yqel1G9Rt67n6RMqYAo6ISCn8vPUgD83YQNJx10c6V3VuxN+Hn0d4tUAvVFaMlN2w+N+w8Su4/Veo08L1mvZXQbMLoN3l0HUMVKvl+TpFyliVCjjayVhEzlZ6Vg7/nB/LB0t2uvTVDA7g78M7MqKrj50XlbgaPrkG0lMcX696F4Y853qdfyCMm+/Z2kTKWZV6qGqtnW2tHR8eHu7tUkSkAok7kMrwKUvchptuzWox756+vhdutv8M/7nqVLgBWPsxZBz3WkkinlSlRnBERErDWstHy3bx/DzXicR+Bu4e1Jq7B7XyvaXfsXNhxljIyXRuD64Jh7dpp2GpEhRwRETcSDqewSNfbWBhMROJX7uuCzFRdbxQ2Wn89jnMnAC2yPEQfR+CAY85HkeJVAEKOCIiRZQ0kXhYF8dE4rAQHwwKK9+BeQ+5tl/yHPS5y/P1iHiRAo6ISJ7snFyenxfL+0t2uPTVDA7gH8PPY3jXxl6o7DSshV8nwY/POrcbP7jyNej2F+/UJeJFCjgiIkBaZjZ3fbbO7SOpbs1q8dp1Xb17qrc7mWmwa4ljCfiGz537/ALhmneg4wjv1CbiZQo4IlLlJZ/IZNyHq1i/O8Wp3c/APYNbc9dAH5xIDDD7Htg4w7U9oBqM+gRa/8nzNYn4CAUcEanSdiencdP7K9medMKpvXGtakwe3YXuzX1wInG+lgNdA05wGFz/JTS/wDs1ifgIH/xfEhERz9i89xhXv7nUJdy0bxjGfyf08X64SUuGbx+DzBPu+6MHOX8d1gRumq1wI4JGcESkiloan8RtH60hNSPbqb1PdARvjenu/VVSWemOjfoObISwRnDhPa7XhDWENkMhtKHjQMxWgyGwmudrFfFBCjgiUuXM2bCXB774jcwc5837rujUkJev7UxwgL+XKitk8cuOcAOw5FWIuRmCQ12vu/4Lz9YlUkFUqUdUxpgrjTHTjh496u1SRMRLPliyg7unr3MJN+MubMHk67p6P9xYC+unw6+vnGpLOwzL3/JeTSIVUJUKODqLSqTqstby4rexTJy9GWud+564rB1PXdEePz/jneLyHU2ET/8MM2+H3KxT7TXqQsw479UlUgHpEZWIVHpZObk8+vUGvlm7x6k9wM/wrz938v5BmdbCmg9hwVOQmeraf/nLUCPC42WJVGQKOCJSqZ3IyOaOT9fyS9whp/bqQf68dWN3+rWp66XK8iTvcOxns+MX176AELjkH9BhmOfrEqngFHBEpFKy1rL/WDq3fbyGDYnO8+4iagTxwc096NSklpeqA3JzYdU78MMzkJXm2t+sDwx7AyKiPV6aSGWggCMilcZPWw/y0dKdxO1LIfnkTk5m5bhc0zyiOv+5uSdRkTW8UGGe/Rth3sOQsMy1L7AGXDwRYv4KflVqmqRImVLAEZFK4a1F8bz4bWyJ15zfOJz3x/agbmiwh6oqxvK33IeblgPgyslQu7mnKxKpdPS/ByJSoRxNyyIt89TmfNZaJn0fd9pw07d1JNPH9/Z+uAEY8Bj4F6ojOAyueh3GzFS4ESkjGsERkQohKyeXJ/+7kS9XJ+JnoE39ULo2q0VWjuWrNYlu7wkJ9KNRrWpc3bUx4/tFExTgI/9PV6sp9LwVlr3h2In4ilccuxWLSJlRwBERn5edk8t9X6xn7oZ9AORaiN2fSux+1yXVNYL8ebhvXUZc2ImwagEY44W9bbIzYPUHsGc1XPOu+2v6Pggt+kHrS8AbNYpUcgo4IuLTcnItD3+1oSDclCQsJIAPx/WketoBwqt74Syp3FzY9BUs/DukJDjautwA0QNdr61eB9oM8Wx9IlWIj4zXioi4ys21PPb1Bv67bs9pr61TI4jp43vTrVltD1TmxrYf4e1+8M2tp8INOJaB5+YWe5uIlA+N4IiIT7LW8rdZm5hRZH5N41rVePemGA6mZrAu4Qib9hwjNCSA+/7UmuYRXlj6nZYM8x+DDcUcepn0ByTFQb12nq1LpIpTwBERn2OtZeLszXy2IsGpvUFYCNNv7U2ziOq0bwj9vb0Lcew8mHMfHD/g2ucX6DgBvN/DULOe52sTqeKqVMAxxlwJXNmqVStvlyIixbDW8vy8LXy4dKdTe93QYKaPd4Qbr0tLhm8fhY1fuu8/byQMehLqtPRsXSJSoEoFHGvtbGB2TEzMrd6uRURcWWv594KtvLN4h1N7RI0gpt/aixbe3H0435Y5MOd+OHHQta9eRxj2OjTu7vm6RMRJlQo4IuLbJv+4jSk/xTu11a4eyKe39qJVvVAvVZUn/SjMfRA2znDt8wtwLPvu+xAEBHm+NhFxoYAjIj5h6s/beOWHOKe2sJAAPv5rL9o1CPNSVYX4B8O+Da7t9c+H4VOgYWfP1yQixdIycRHxuncXb+el+Vud2kKDHeHmvMbhXqqqiMAQGP4mmLx/Nv0CYMDjcOtChRsRH6QRHBHxqo+W7eQfc7c4tdUI8ufDcT3o3LSWd4oqTpPu0Ocex543w6dCw07erkhEiqERHBHxms9WJPB/s353aqsW6M/7Y3vQvXkd7xR14jDEfVd8/8AnYPxPCjciPk4BR0S8Ysbq3Tw5c6NTW3CAH+/eFEOvlhHeKWrnEpjaG778i2ODPncCgsHfC8dAiEipKOCIiEfl5lreXbydR77egLWn2oP8/Zj2lxgubBXp+aKsheVvwn+udCz/zk6HWXdCbo7naxGRMqE5OCLiMYdSM3hoxm8sijvk1B7ob3jzxm7e2Zk4Mw1m3+O6/Hv3Cvjtc+h6g+drEpFzpoAjIh7x09aDPDzjN5KOZzq1+/sZXh/djcHt63u+qOQd8MUYOLDRta/fw9DpWs/XJCJlQgFHRMpVelYO/5wfywdLdrr0hYYE8PKfO3NJxwaeL+yPH+Drv0J6inN7cBhcPQ3aXur5mkSkzCjgiEi5+eNAKvd8vp4t+4659PWIqs0ro7rQpLaHz5bKzYVfJ8HCfwDWua9uOxj1KUTqvDqRik4BR0TKnLWWz1Ym8Pc5m0nPynXq8zNw7+A23DkwmgB/D69zSD8GM++A2DmufR2GwbCpEFzTszWJSLlQwBGRMnXkRCaPfr2BBZsPuPQ1rlWN167rQkyUF/a4ObQVPr8BDhdZ/m38YPDTcOG9YIzn6xKRcqGAIyJlZum2JO7/cj0HjmW49F3ZuRH/GH4e4dW8sIfMwVh4dzBkHndur1YHRr4P0QM9X5OIlCsFHBE5Z1k5uUz6Po63FsU77W0DjmMXnh12Hld3a4zx1ghJZBtofiH8UWiH4gadYNQnULu5d2oSkXKlgCMi52Rn0gnu/XwdvyUedenr1CScydd1JSqyhhcqK8TPz7Ey6p2BkLwdOo+GK16BwGrerUtEyk2VCjjGmCuBK1u10goJkXNlreXrtXt4etYmTmQ67/hrDNzeP5r7/9SGoAAPTyQuOoSUr1otxwqpXUugxy2abyNSyVWpgGOtnQ3MjomJudXbtYhUZEdPZvG3mZuY/dtel776YcG8cm0X+nj6yIXcXFj9Hmz6Bnr/y/019Ts4PkSk0qtSAUdEzt2aXcncM309e1JOuvRd0qE+/7ymE7VrBHm2qKN7YNYE2P4zAOH1ZkOH8zxbg4j4FAUcETkj2Tm5TPkpntd+jCO3yFOgkEA/nrqiA9f3bOb5icSbvoY590P6qTlAdTe+DX+6E0LCPFuLiPgMBRwROa3EI2nc/8V6Vu084tLXrkEor4/uSuv6oZ4vbMFTsHSyS7PJyYD9GyDqIs/XJCI+QQFHREo0Z8NeHv9mI6np2S594y5swSND2xIS6O/5wla/7zbcENWXHec/RCuFG5EqTQFHRNw6kZHNxNm/8+XqRJe+yJpB/OvPnRnYtp4XKgN2LIZ5Dzu3+Qc5diTuPYHsuDjv1CUiPkMBR0RcbEw8yj2fr2NH0gmXvv5t6vLvP3embmiwFyoDknfAl3+B3EIjSoHVYewcaNzdOzWJiM9RwBGRArm5lncWb+ffC7aSleM8kzjI349HL23HzX2i8PPz0h4y6cdg+mg4mezcPuJthRsRcaKAIyIAHDiWzgNfrmfJtsMufdF1azB5dFc6Ngr3QmV5Dmx2rJY6tMW5feCT0OEq79QkIj5LAUdE+H7zAR756jeOpGW59F3fqxlPXd6BakFemEicb9M38PVfweY6t3e8Gvo97P4eEanSFHBEqrD0rByem7uFj5fvcumrVT2QF6/uxNDzGnihsiJaDoDgMEhPOdXWsAsMm6IjF0TELQ8fEiMivmLbweNc9cavbsNN75Z1+Pbevr4RbgCq14GBT5z6ukV/uGEGBFX3Xk0i4tM0giNSBc3ZsJdHv9rgckhmgJ/hgUvacFu/aPy9MZH40FaIbON+VCZmHMQvhG43QdtLNXIjIiVSwBGpQrJycnlhXizvL9nh0tesTnUmj+5Kl6a1PF/Y8YOw8O+w9mO45l04f6TrNf6BcP0Xnq9NRCokBRyRKmL/0XTu+mwtq3e5HrcwvEsj/j78PEJDAj1bVFY6LJ8KiydBZqqj7funod3lEFjNs7WISKWigCNSBSyNT+Ke6etIOp7p1B7ob/i/KztyYy8PH5JpLWyeBd8/BSkJzn3HEmHpG9Bfq6NE5Owp4IhUYtZa3v5lOy/Nj3U5AbxReAhTb+zu+UdSWSdh9r2woZjHTRGtoVFXz9YkIpWOAo5IJXUsPYuHvvyNBZsPuPT1bR3Ja9d1pU6NIM8WlZIAn9/gOOm7qJBaMOBx6PFXx3wbEZFzoIAjUglt2XeMOz5Zw87DaS599wxuzb2DW3t+ldSOX2DGWEgrslOyXwD0uAX6P+pYDi4iUgYUcEQqma/XJPLkzI2kZznv+hteLZBXR3VhYDsPnwBuLax4C757EqzzsnQi28K1/4F67T1bk4hUego4IpVERnYOE2dv5rMVCS595zcOZ+oN3Whax8Mb42WdhNn3wYbPXfvaXQEj3oLgUM/WJCJVggKOSCWQeCSNCZ+uZUPiUZe+0T2b8fSVHQgJ9MJZUgnL3YQb4zggs++D4KfN1EWkfCjgiFRwP289yH1frCelyEGZwQF+/GP4efw5pqmXKgOiBzrm1iz6Z15RYXD1O9B2qPdqEpEqQQFHpILKzbVMXvgHr/34B7bIEvDmEdWZekM3OjYK905xhfV/DPZtgOR4uO4ziGzt7YpEpApQwBGpgI6cyOS+L9azKO6QS9+f2tfn5Ws7E17NR5Za+/nB1dMcfw4J824tIlJlVKkH4MaYK40x044edZ2nIFJR/LY7hSte/9Ul3PgZeHRoO6aN6e75cHM0Eb7/P8jNdd8fEqZwIyIeVaVGcKy1s4HZMTExt3q7FpHSstby2coEJv5vM5k5zkEismYQk0d3pU90pOcL2/krfHkTpCVBQAgMfMLzNYiIFFGlAo5IRXUyM4cnZ27km7V7XPq6N6/NlOu70SA8xLNFWQsr3obvnji1v82if0KDTtD+Cs/WIiJShAKOiI/bkXSCOz5ZQ+z+VJe+my+M4onL2hPo7+GnzVknYc798Nt0174t/1PAERGvU8AR8WHf/b6fh778jdSMbKf26kH+/POaTlzZuZHnizqa6DhPat96174Bj0O/Rzxfk4hIEQo4Ij4oOyeXfy3YytuLtrv0RdetwdtjutOqnhd2AC4836awoFDHSql2l3m+JhERNxRwRHzModQM7p6+luXbk136rujUkBev6UTNYA//p2strJzmmG+T6zyaRERrx/42ddt4tiYRkRIo4Ij4kNU7k5nw6VoOpmY4tQf4GZ68vD1j+0RhjIdPAc9Kz5tv85lrX5tL4eq3IcQHNhQUESlEAUfEB1hreX/JTl6Yt4XsXOdtieuHBTP1hm50b17H84UdTYQvboS961z7+j/mOIZB50mJiA9SwBHxsuMZ2Tz61Qbmbtzn0ndBywhev74rkTWDPV9Ybg58fDUkbXVuDwp1jNq0u9zzNYmInCEFHBEv+uNAKrd9sobth0649N0xIJoHL25DgKeXgOfz84dL/wmfXA02b2PBiFZw3XTNtxERn6exZREv+d9vexk2ZYlLuAkNCWDamO48OrSd98JNvuiBcPGzjj+3GQq3LlS4EZEKQSM4Ih6WmZ3L8/O28OHSnS597RqE8taN3YmKrOH5wopzwV0Q3gTaD9N8GxGpMBRwRDxo39GT3PnpWtYmpLj0XdOtCf8Yfh7Vgvw9X1jCcsfjpxpuzrIyBjqO8HxNIiLnQAFHxEOWbEvinunrOHwi06k9yN+PicM6cl2Ppp5fAm4trHoX5j8GzS6AMTPBX/8siEjFp3/JRMpZbq7lzUXxvLxgK0VWgNO4VjXevLEbnZrU8nxhWekw70FY94nj652L4funYOgLnq9FRKSMKeCIlKOjJ7N48Mv1/LDloEtf/zZ1eXVUF2rXCPJ8YSkJ8MUY1/Oklk+FLjdAg/M8X5OISBlSwBEpJ7/vPcodn6wlITnNqd0YuG9wG+4e1Ao/Pw8/kgLY9iN8/Vc4ecS5PagmjHhb4UZEKgUFHJFyMGP1bv42cxMZ2blO7bWqB/LadV3p36au54vKzYVfX4aFzwFFnpXViXacJ1WvnefrEhEpBwo4ImUoPSuHibN/Z/rK3S59nZuEM+WGbjSpXd3zhZ1Mgf/eDnHfuva1vRxGvKnzpESkUlHAESkju5PTuOPTNWzac8yl78bezXjqig4EB3hhCfiB3x3nSSVvd243fjDoKbjwPu1vIyKVjgKOSBn4KfYg932xnqMns5zaQwL9eH7E+VzdrYl3CtswA/53N2SfdG6vVgdGvu/YqVhEpBJSwBE5Bzm5ltd+iGPywm0ufVER1XlrTHfaNQjzQmXAyndg3kOu7Y26wrUfQ62mnq9JRMRDNC4tcpaST2Qy9oOVbsPNkI71+d/dF3kv3AC0vwpqNnBu6z4Wbp6vcCMilZ5GcETOwrqEI9z56Vr2Hk13avf3Mzw6tC239m3p+V2JiwqtD9d+BB9eBsYfLn8Zuo3xbk0iIh6igCNSCtZaPlm+i2fnbCYrx3mpdWTNYN64viu9W0Z4qTo3mvWCq96Aeu2hURdvVyMi4jEKOCJnKC0zmye+2cjM9Xtd+npE1WbK9d2oFxbi+cIyjsORHdDgfPf9XUZ7th4RER+ggCNyBrYfOs4dn6xl64FUl75b+7bgkaHtCPT3wpS2pD8cS8BPJMFtv0B4Y8/XICLigxRwRE5j/qZ9PDRjA8czsp3aawT5868/d+ay8xt6p7Ats+G/d0BmXuiacROMnQcBXjjbSkTExyjgiBQjOyeXl77byrRftrv0ta5Xk7fGdCe6bk3PF5aTDQv/DktedW5PXAVLJ0M/N0vDRUSqGAUcETcOHkvnrunrWLkj2aXvqs6NeOHq86kR7IX/fI4fgq/HwY5fXPvOvxZ6T/B8TSIiPkgBR6SIFdsPc9f0dRxKzXBqD/Q3PHVFB8b0bu6dJeCJq+HLv8CxPc7tfgEw9EXocYvjqHIREVHAEclnreXdxTt4cX4sObnOS8Abhocw5YZudGtW2xuFwZoP4NtHISfTuS+0Ifz5P47l4CIiUkABRwRITc/ika828O2m/S59F7WK5LXruhBRM9jzhWWdhLkPwvpPXfuaXwgjP3Bs6CciIk4UcKTK27o/lTs+WcP2pBMufXcNbMX9F7fB388Lj36O7HQsAd+/0bXvgrvgT8+Af6CHixIRqRgqbMAxxvQDHgK6A42Am621H3q1KKlwZq7bw+PfbORkVo5Te1hIAK+M6sLg9l4aHTkUB+9dDOkpzu2BNWD4FOg4wjt1iYhUEBU24AA1gU3AR3kfImcsIzuH5+Zu4aNlu1z6OjYK480butMsoroXKssTEe049Xv7T4XaWsOoT6BeO+/VJSJSQVTYgGOtnQfMAzDGfOjdaqQi2ZtykgmfrmX97hSXvlExTZk4rCMhgf5eqKwQP3+45j2Y1h+O7ob2V8KwqRDixdPJRUQqkDMOOMaYkUB/oAvQGQgFPrXW3ljCPU2AZ4GhQASwD5gJTLTWHjmHukXOyq9/JHHP5+tIPuG8GikowI+/D+vIqB7NvFSZGzUiHKeB71wMfe7REnARkVIozQjO33AEm+NAIlDiOLkxJhpYCtQDZgGxQE/gXmCoMeZCa+3hsylapLRycy1Tf97Gy9/HYZ1XgNO0TjXevKE75zUO905xB36H+h3d9zXu5vgQEZFSKc3pgPcDbYAw4I4zuH4qjnBzj7V2uLX2MWvtIOAVoC3wXOGLjTH/MMbY03wMKEW9IgAcz8jmjk/X8O8FruFmULt6zLmrr3fCTXamYwn4mxfCHz94/vuLiFRiZzyCY60tmO14ul1c80ZvLgF2AlOKdD8NjAfGGGMetNbmr819FfjkNGUknGm9IgAJh9O49aPVLqeA+xl48JK23NE/Gj9vLAE/usdxOGbiKsfXX/8VblsEtaM8X4uISCVUXpOMB+Z9XmCtzS3cYa1NNcYswRGAegM/5rUnAUnlVI9UQUu2JXHnZ2tJSctyaq9TI4jJ13XlotaR3ilsxy8w42ZIK/R2T0+Bb26DcfM110ZEpAyUV8Bpm/c5rpj+P3AEnDbkBZzSMsbUBFrlfekHNDPGdAGSrbUuIz3GmPE4Ro5o1KgRsbGxZ/Ntq5ykpKQK93dlrWXWlqNMW3WYIicuEF0niKcHNSAyJ4nYWA/naWupE/sJdTdMxTjnfjJrNGJPhzvJ2LrVszVVUmaJ9AIAABxiSURBVBXxfStVm96zZa+8Ak7+hIajxfTnt9c6h+8RAxTaJISJeR//AcYWvdhaOw2YBhATE2PbtdNeImciNjaWivR3lZGdw5P/3cRXa1znr1/RqSH/GtmZakFeWAKekQozJ8CW/7n2tb6EoKun0aKaF865qqQq2vtWRO/ZsleR98H5GdBYvhQ4eCyd2z5Zw7oE5/1tjIGHhzjm23jlFPBDWx1HLiQVHdA0MOAx6PcI+JVmvr+IiJxOeQWc/BGa4pam5Le77rQmchbW707hto9Xc+BYhlN7aHAAr43uwqB2Xjpy4ff/wqy7IPO4c3tIOFz9LrS5xDt1iYhUcuUVcPInErQppr913ufi5uiInLGv1yTy+H83kpntPK+lZWQNpv0lhlb1anq+qJxs+OFpWPaGa1+D8+Haj6FOC8/XJSJSRZRXwMmfG3OJMcav8EoqY0wocCGQBiwvp+8vVUB2Ti4vfhvLu7/ucOnr36Yuk0d3JbyaF07bzs2BT0Y4VksV1fl6uPxlCPLiOVciIlVAuTz4t9bGAwuAKODOIt0TgRrAx4X2wBEplZS0TG7+cJXbcHNb/5a8P7aHd8INOM6RatGvSFsgXD4Jhk9VuBER8YDSnEU1HBie92WDvM8XFDroMsla+1ChWybgOKphsjFmMLAF6IVjj5w44MlzqPusGGOuBK5s1arVaa8V3xV3IJVbP1rNrsNpTu3BAX7885pODO/a2EuVFXLRg5C4BuK+hdBGMOpjaBLj7apERKqM0jyi6gLcVKStZd4HwC6gIOBYa+ONMTGcOmzzMhyHbb6Glw7btNbOBmbHxMTc6unvLWVjwe/7uf+L9ZzI/P/27jw+6ure//jrJKyyBBBkUVELIlqRtW611n2nioqIXqitRYta2tqft7X2Wu21tYvaWqstUK1eUbRaq6LXSq1roSoiWjdQQUCRKotEZBGSnN8fE26ZTAIJJN/vZOb1fDzyGHPOd2Y+efB18s53PnNOZdZ4z7I2TBwzlP122Z6VBxpRSQmM+B08ehkcdQW075Z2RZJUVBqyVcMVwBUNefAY47vAVxpWkpQrxshvHs9sllnT0N0689v/GMJOHdokX9iqd6Fjr8zbUjW17QSn1NypRJKUBBffUN5b82kFF975Yq3h5szP7cqd4w5IJ9zM+0tmo8zHr0r+uSVJW9RsF/pTcXh3ZWazzLn/yt4ss7Qk8MPh+zDmwN2SX7yvqhKe/Ck8/fPM93+/DnYeCnuflGwdkqQ6GXCUt/4xfwUX3DGbj2psltl5h5bcdPZQDuqzY/JFrV0J942Dtx/LHr9/PPQaDGV50OAsSTLgKP/EGLn92UVcOe11Kmvsltm/Rwcmjx3Grl1S+Kj1+y/BH8fAqhp7uYbSzJYLHXslX5MkqVYGHOWVDRVVXP7Aq9w1692cueP37cE1IwfSrnUKp+2cKfDwd6BiffZ4u24w8lbY/ZDka5Ik1amoAo7r4OS3Zas/ZfyU2bywKHcFge8c3Y+LjuibfL9NxafwyH/C7Ftz53Y9IBNuvHIjSXmnqAKO6+Dkr3++t4rzb5/N0vLsKyTtWpXyy1GDOOazPeq4ZxMqfw/uHgPvv5g7t//5cMxV0KJV8nVJkraqqAKO8tP9c5bw3T/9k09rbJa52447MHnsMPp175B8UQuehHu/CmtXZI+3aAvDr4eBo5KvSZJUbwYcpaayKvLzv8xl4tMLcua+sGdXbhg9mE47pHCF5PnJmbelYnbgovMeMGoK9Ng3+ZokSQ1iwFEqytdtZMLUOTz15rKcuXMP2YNLj+9Pi9KU1qHcaW+gRq9Pv+MzWy+0zZOtICRJW+RKxkrc2x9+wik3zsgJN61alHDNyIH810n7pBduIPOJqKN/VP1NgCN+AGfeabiRpGbEKzhK1ONzP+CbU19i9acVWeM7dWjNxDFDGdy7c0qV1XDQhbByPvQ/EfoelXY1kqQGMuAoETFGfvvUfH7x6Dxi9tp9DNq1ExPHDKV7x4T3k6rcCJ+uhh265M6FACf9Mtl6JEmNpqgCjuvgpGPdhkouufdlHvrn0py504bswo9H7EublrXsxt2UVn8A95yT+e8vPwilLZN9fklSkyqqHpwY47QY43llZWVpl1I0lqxax+m/m5kTbkpLApeftA/XjNwv+XCz6B8w8VBYPDPz9dfLk31+SVKTK6orOErW8++sZPyU2axYsyFrvKxtS248awiH7Nk12YJihOcmwvTLoGqzHqBnb4LeB8I+JydbjySpyRhw1CTueG4RP3zgNSpqbJbZr3t7Jo8dxm47tku2oA1rYdoEeOWe3LnPHAa7uZeUJBUSA44a1YaKKq6c9hp3PLc4Z+6Yfbpz3ahBtE96s8zy92DqaPjXP3PnDrk48zHwkoTfJpMkNSkDjhrN8k8+5YIpL/L8wpU5cxOO3JNvHbknJSUJb5b57iy46yxY82H2eOuOmYX7+p+YbD2SpEQYcNQoXl1Szvm3z2bJqnVZ4zu0KuXakQM5fkDP5It6aWrmbanK7B4guu2d2XKhq5+mk6RCZcDRdpv28vtccu/LrN+YvXfTLp3bMnnsMPbu2THZgqoq4bErYOavc+f2OgFOnQStU9jAU5KUGAOOtllVVeSa6fO46cn5OXMHfWZHbjx7CF3aJbxZ5vqP4U9fg7cezZ075GI44r+gpKhWR5CkomTA0Tb5eP1Gvn3XS/xt7oc5c+ccvDuXnbg3LdPYT+rTj+H9Odljpa3h5Bthv5HJ1yNJSkVR/SkbQhgeQphUXl6edinN2oJlnzDixhk54aZlaeBnpw3gii99Np1wA1C2C5x5B5RWXzlq3wO+8ojhRpKKTFEFHFcy3n5PzvuQk2+cwfxla7LGu7ZvzV3nHcioz/VOqbLN7Lo/DP819BoM5z0BuwxNuyJJUsJ8i0r1EmNk0tML+Nlf5lJj7T7226WMiWOG0rOsbTrF1WbQaBgwEko9xSWpGBXVFRxtm/UbK/n23S9x9SO54WbE4J354/kHJR9u1q7MNBOv/lfdxxhuJKlo+RtAW7RsTQWX/O4fvLIku2+pJMD3ju/PuC98hhASXrzvw7kw9Uz46B1YuQDOeRha5tHVI0lS6ryCozrNXrSSCdPeywk3Hdq04JZzPsd5h/ZJPty8OR1+f1Qm3AAsmQ0PTshspClJUjWv4KhWf5z1Lpfd/wobK7ODQ59u7Zg8dhif6dY+2YJihJk3wF8vB2qEmfJ3YcMaaJ1wTZKkvGXAUZbKqsjV//sGv//7OzlzR/bfiV+dOYgObVomW9TG9fDQt+Dlqblzg8fAiddBi4QXFJQk5TUDjv7P6vUbmTB1Dk/MW5Yzd+Hhfbj46L0oTXqzzNUfwN1nw3uzssdDCRz7Ezjg65D022SSpLxnwBEAi1es5Wv/M4s3P/gka7xVaeDaMwYxfGCv5Ita+jJMHQ0fL8keb10GI2+BvkclX5MkqVkw4IjnFqzg61Nm89HajVnj3Tu25rJDu6YTbl77M/x5PFRk707Ojn1h9F3Qdc/ka5IkNRtFFXBCCMOB4X379k27lLxx96zF/OD+V3OaiffbpYxJY4axaunCZAuqqoKnfgZP/TR3rs8RcPot0LZzsjVJkpqdovqYuFs1/FtlVeS/H3qd7/4p95NSJ+7Xk7vPO4geZW2SL+yZa2oPNwdeAGfdY7iRJNVLUV3BUcaWmom/fVQ/JhzZN/n1bTYZ9lWYczusWpz5vqQlnHQdDBmbTj2SpGbJgFNkFq9Yy7m3zeKtD7Obidu0LOHakYM4cb+eKVVWrV3XTI/NzcdAi9YwagrsdnC6NUmSmh0DThF5dsEKxtfRTPz7sZ9jwC558tZd989mgk2Xz0Dn3dKuRpLUDBlwisRdz2eaiStq7JY5cJcyJo0dRveOCffbVFXCynegax0N330OT7YeSVJBMeAUuMqqyI8ffoNbZuSuTHzSfj25ZuRA2rQsTbao9eWZncDfmwXjnoAueyT7/JKkgmfAKWAfVzcTP1lLM/HFR/fjG0ek0Ey8Yn5m8b7l8zLfTx0N506HNh2TrUOSVNCK6mPixWTRijWcetPMnHDTpmUJN509hAlH7pl8uFnwFEw+4t/hBmDZG/DopcnWIUkqeF7BKUDPVq9MvKpGM3GPjm2YPHZYOs3Ez0+GR74LsTJ7vNcQOPwHydcjSSpoBpwCM/X5xfxXHc3Ek8cOY6ekm4krN8Ij/wkv3JI7t+/pcPJvoGXbZGuSJBU8A06BqKis4if/O7fWZuLhA3vxi9P3S76ZeO1K+ONYWPhM7tyRl8MhF7sTuCSpSRhwCsDH6zfyjTvn8NSbedRM/OEbMPVM+Ghh9njLdnDaZOh/YrL1SJKKigGnmVu0Yg3n3vYCb9eyMvEvzxjE8QNSWJl43l8yHwPfsDp7vFPvzCrF3T+bfE2SpKJSVAGn0HYT/8f8FYy/o/Zm4t9/eRj77pxwM3GMMON6eOwKILsHiN4Hw6jbM1sxSJLUxIrqY+KFtJv41OcXM+bm53LCzcBdO/HgRZ9PPtwAxCpYNJOccDNkLIx9wHAjSUpMUQWcQlBRWcWV017j0vteyfmk1PCBvbj7vAOT/6TUJiWlcNrvoVv/zPehBI77GQz/NbRolU5NkqSiVFRvUTV3H6/fyEV3zuHpWpqJv3N0Py5Ko5m4pjYdM302U06FE6+FPkekW48kqSgZcJqJhcvXcO5ts5i/bE3WeNuWpVx3xsB0monr0mUPuHAWlHp6SZLS4W+gZmDm/OVccMeLOf02PcsyKxMn3m9TVQVP/RR6DoL+J9R+jOFGkpQifwvluTufW8zlD+SuTDxo105MGjM0+X6bDWvgz+fDG9OgVfvMRpl+7FuSlGdsMs5TFZVVXPHga3z/z7nNxCcP6sVdaTQTr1oMNx+bCTcAGz7JLOa3ZkWydUiStBVewclD5es28o2ptTcTX3LsXlxwWJ/km4kXPwt3/wesqVHThrVQvhja7ZhsPZIkbYEBJ89sqZn4l6MGcty+KTQTz5kC074FVdk9QHTfF0ZPzaxQLElSHjHg5JGZ85czfsqLlK/Ll2biSnaacz3MuzN3rv9JMGIitG6fbE2SJNWDASdP3PHcIn74wGu1NxOPHcpOHRLut1lfDvd+lS5vP5Y7d+glcNj3ocQWLklSfjLgpKyisoqrHn6DW2cuzJk7ZVAvfnrafrRpWZpsUSvmZ5qHl7+ZPd6iDZx8Iww4Pdl6JElqIANOisrXbeSiO1/kmbeW58yl1kw8/wm45xxYvyp7vENPOPNO2HlIsvVIkrQNDDgpeae6mXhBrc3Egzhu3x7JF/X+SzDlNIiV2eM7D4VRd0DHPFotWZKkLbCJIgUz317OKTfOyAk3vcracO/4g9IJNwA9B+a8/VS+23FwzsOGG0lSs+IVnITlXTPx5kLI7Py94m1Y8iIceTlLux5PWcu26dUkSdI2MOAkJC+biWvTsk2m12bpy9DvWJg7N+2KJElqMANOAvKymXjZPOi2V+1zHXpkviRJaqaKqgcnhDA8hDCpvLw8sed8Z/kaRtw0Iyfc7NCqlIljhnLh4X2TDTcxwjPXwo0HwEtTk3teSZISVFQBJ8Y4LcZ4XllZMisCb7GZ+OsHc+xnE75KsnEd3DcO/vYjIMK0CfDurGRrkCQpAUUVcJI05dlFjLnl+ZxtFwb37sT9F32efXp1TLagj5fCH06AV+7591jlBrjny7BxfbK1SJLUxOzBaWQVlVX890Ovc9s/FuXMjRi8M1efOiD5ZuIlL8JdZ8HqpdnjbTrBKTdlGoslSSogBpxGVL52IxdNzW0mDiHTTDz+iyk0E79yLzxwIVTUuErTtR+Mvgt27JNsPZIkJcCA00gWLPuEr932AguWZ/fb7NCqlF+NGsQxSffbVFXBE1dlGopr6ns0nH4ztEl4d3JJkhJiwGkEM95ezvgps/l4fUXW+M6d2jJ57LDk+20+XQ33nQ/zHs6dO/gbcNSVUJIHa+5IktREDDjbKcbIbx5/OyfcDOndiYljhtGtQ+tkC/poEUwdDR++lj1e2gqGXw+Dzkq2HkmSUuCnqLZTCIEbzhrMzp3+vZ3BqYN35s5xByYfbhbOgMmH54abdt3gyw8ZbiRJRcOA0wi6tm/NzecMo0PrFnz3uP5ce8bAdLZdWD4P1q7IHusxAMY9Ab0PSL4eSZJS4ltUjaR/j448eclh7Ng+4as2mxv2VfjgdZg1OfP93l+CEb+DVu3Sq0mSpBQYcBpRquFmk+OuhhVvwa4Hwhe/CyVepJMkFR8DTqEpbQln/wlK/aeVJBUv/7xvjt5+DB69LLNxZm0MN5KkIudvwuYkRnj2tzD9MohV0Kk3HHB+2lVJkpR3vILTXFR8Cg9eBI9emgk3AH/5Hsx/PN26JEnKQwac5uCTZXDbl2DOlNy5FfOTr0eSpDznW1T57l+vZFYmLn83e7x1Rzj9Ftjz6HTqkiQpjxlw8tkb0+C+82Dj2uzxznvAWXdDt73SqUuSpDxnwMlHMcLTv4Anfpw7t8cXYeStsEOXxMuSJKm5MODkmw1r4YEL4bX7cuf2Pw+O/UlmrRtJklQnA04+KV8Cd50FS1/KHi9pASf8IrMVgyRJ2ioDTr6o2AC3ngAfLcweb9sFzvgf2OMLqZQlSVJz5MfE80WLVnDk5dlj3faGcY8bbiRJaiADTj7Z9zQ49JLMf/c7Hs6dDl32SLcmSZKaoaJ6iyqEMBwY3rdv37RLqdth34cd+8KAM9wJXJKkbVRUv0FjjNNijOeVlZWlW8hHi2DDmtrnSkpg4JmGG0mStoO/RZP2zjMw6TC4fzxUVaVdjSRJBcmAk6RZN8Ptp8C6lfD6A/D0z9OuSJKkgmTASULlRnj4O/DwxVBV8e/xJ6+Gxc+lV5ckSQWqqJqMU7F2JdzzZXjn6dy5I34Au+6ffE2SJBU4A05T+nAuTD0TPnone7xlOzh1Iuw9PJ26JEkqcAacpvLmdLj3q7BhdfZ4WW8YPRV67JtOXZIkFQEDTmOLEWbeAH+9HIjZc70PgjNuh/bdUilNkqRiYcBpTBvXw0Pfgpen5s4NHgMnXpfZkkGSJDUpA05jWf0B3H02vDcrezyUwLE/gQO+DiGkU5skSUXGgNNY1pfDsjezx1qXwcg/QN8j06lJkqQi5To4jaVbPzj9lswVG8jsJzXub4YbSZJS4BWcxrTnUXDMVfD2Y5mw07Zz2hVJklSUDDiN7cALMv02JaVpVyJJUtEy4DS2ECAYbiRJSpM9OJIkqeAYcCRJUsEx4EiSpIJjwJEkSQXHgCNJkgqOAUeSJBUcA44kSSo4BhxJklRwDDiSJKngGHAkSVLBMeBIkqSCY8CRJEkFJ8QY064hcSGEZcCiWqbKgPLtfPhtfYyG3K++x9bnuK0d0xVYXs+6moPG+DfOt+fe3sdN4pxtyPGet7nSOm/z9ZzdnsfwtTY5TXX+7BZj7LbVo2KMflV/AZPSeoyG3K++x9bnuK0dA7yQ9r9Lvv0b59tzb+/jJnHONuR4z9vkzp20ntfX2lrnC+qcbax/5+358i2qbNNSfIyG3K++x9bnuMb4mZuTNH/epnru7X3cJM7ZhhzveZsrrZ83X8/Z7XkMX2uTk+rPXJRvUan+QggvxBiHpV2H1BCet2puPGcbn1dwtDWT0i5A2gaet2puPGcbmVdwJElSwfEKjiRJKjgGHEmSVHAMOGpSIYRDQwgPhhCWhBBiCOGctGuSNgkhXBBCeCeEsD6EMDuE8IW0a5K2xNfU+jPgqKm1B14FvgmsS7kW6f+EEEYB1wM/AQYDM4FHQgi9Uy1M2jJfU+vJJmMlJoTwCXBRjPHWtGuRQgjPAf+MMY7bbOwt4N4Y46XpVSbVj6+pW+YVnAITQjg9hHBDCOGZEMLH1Zcwp2zlPruEEG4JIbwfQvg0hLAwhPCrEELnpOqWamrKczmE0AoYCkyv8RDTgYMb9ydRMfE1OH+0SLsANbofAAOBT4D3gP5bOjiE0IfMpfmdgAeAucD+ZC5/HhdC+HyMcUWTVizVrinP5a5AKfBBjYf5ADiqsX4AFSVfg/OEV3AKz7eBfkBHYHw9jr+JzP9YE2KMp8QYvxdjPAL4JbAX8OPNDw4hXFX9F8mWvg5r3B9JRapJz2WpiXje5gkDToGJMT4RY3wr1qO5qvovh2OAhcCNNaZ/CKwBxoQQ2m02/itg7618Pb+dP4bU1OfycqAS6F7j2O7Av7anbhW3BF6DVU8GnOJ2ePXt9Bhj1eYTMcbVwAxgB+DAzcaXxxjnbuVrbXI/ggQ08FyOMW4AZgNH13ico8m8XSAlocGvwao/A05x26v69s065t+qvu23rU8QQmgfQhgUQhhE5nzrXf29H8VVY9qWc/k64JwQwtdCCHuHEK4HegG/a6IapZoafN76mlp/BpziVlZ9W17H/KbxTtvxHMOAOdVfbYErq//7R9vxmFJNDT6XY4x3A98i0xT6EnAIcEKMcVFTFSnVsC2vwb6m1pOfolKTijE+CYS065BqE2O8iUyTp9Qs+Jpaf17BKW6b/jooq2N+0/iqBGqRtofnspojz9smZMApbvOqb+vqsdmz+rau94elfOG5rObI87YJGXCK2xPVt8eEELLOhRBCB+DzwFrg2aQLkxrIc1nNkedtEzLgFLEY43wyS9PvDlxYY/pKoB1we4xxTcKlSQ3iuazmyPO2abnZZoEJIZwCnFL9bQ/gWGAB8Ez12PIY4//b7Piay4S/ARxAZn2GN4GDXSZcafBcVnPkeZs/DDgFJoRwBZkVMOuyKMa4e4377ErmI4bHATsCS4E/A1fGGD9qmkqlLfNcVnPkeZs/DDiSJKng2IMjSZIKjgFHkiQVHAOOJEkqOAYcSZJUcAw4kiSp4BhwJElSwTHgSJKkgmPAkSRJBceAI6lZCiEsDCEsTLsOSfnJgCOp4IQQRoYQYgjhtLRrkZQOA46kQjQCWAc8knYhktJhwJFUUEIIrYATgekxxrVp1yMpHQYcSXkrZFwUQngthLA+hLAkhPCbEELZFu52BNCRzG7Mmx7n1uq3rHYPIZwfQnil+vE+CCFM2srjSWqGWqRdgCRtwa+ACcBSYBKwETgZOABoBWyo5T6nAhXAtFrmfg4cWz03HTgcGAf0JROMJBUIA46kvBRCOJhMuJkP7B9jXFk9fhnwBNATWFTjPiVkAtBTm46v4UBgQIxxcfXxLYDHgcNDCPvHGJ9vqp9HUrJ8i0pSvvpK9e2PNw8rMcb1wKV13OdgYCc2e3uqhh9tCjfVj1UB/KH62/23r1xJ+cSAIylfDam+faqWub8DlbWMjwAicH8dj/lCLWPvVt92blB1kvKaAUdSvtrU+PtBzYnqKy/La7nPCGBWjHFJHY+5qpaxiurb0gZXKClvGXAk5avy6tvuNSeqe2e61hgbBOwB3Nf0pUnKdwYcSfnqxerbL9Yydwi5V1xGVN/W1X8jqYgYcCTlq1urby8LIXTZNBhCaANcXcvxI4DXY4xvJlCbpDxnwJGUl2KMM4AbgD7AqyGEX4cQrgVeJbPExdJNx4YQ+gID8OqNpGoGHEn57JvAN8j045wPjAYeBY4ie5G/TW9P2X8jCYAQY0y7BknaLiGEmUCvGOPuadciKT94BUdSsxZC6ElmheK61r6RVIS8giNJkgqOV3AkSVLBMeBIkqSCY8CRJEkFx4AjSZIKjgFHkiQVHAOOJEkqOAYcSZJUcP4/QT/djpgitTkAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 576x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAjEAAAGoCAYAAAC37rTiAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4xLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvDW2N/gAAIABJREFUeJzsnXd4VcX2sN9JIYEkpJBCQkkoQUAUIRAQqVIVCwKCgEBAAnYQ/Xm91wLYrn5XQbGTAEGkqFRFAQMmUqQrIkjoCSShBQiEkJ75/tjnHE7PSSMJzPs8+zlnz6yZvfbstvaaNbOFlBKFQqFQKBSKmoZTVSugUCgUCoVCURaUEaNQKBQKhaJGoowYhUKhUCgUNRJlxCgUCoVCoaiRKCNGoVAoFApFjUQZMQqFQqFQKGokyohRVBlCiDghhNQtyVWtT3VACHGvEGKDEOKSEKLYqH3Cqlo3RdVgdA5IIcT0KtIh0UiHxDKUjzLbj7AKV7ICEEK0E0IU6XTcXIpyPc32r2clqlltEUK8bdQGk2/ENpURUwqEEH5CiJeEEOuFEOlCiFwhRI4Q4rgQYrEQYrAQonZV66momQghWgM/Ab0BH0CUomyY2U1UCiHOCCHcbMjvtSIfVkJ9hUKIbCFEmhBihxBijs7osqqnEGK6urHf/NxkLyMfcv25+N+qVMQYIYSrEGKKEOJrIcR+3bXocJsLIToLIb7VXbt5QogMIUS8EGKkreu3jOU+BnJ0/18XQviUbY8dx6WyN3CzIIQYC3wCeFnJbqJbRgDjgLgbp1mNZimwX/f/clUqUk0YCrjr/kvgMyBFt36xDPUFAaOAecaJQog+QNsy1OcM1NEtIUAkEA1sF0KMkFIml6FORc3gC2CN7v+pMpTfBfyf0XpZzudKRQjRF+ilWz0kpfy5KvUxwwOYVZaCQoiXgP+ZJdcD+uiWIUKIx6SUBeUtJ6U8L4RYCEzUyb4IvF4WvR1FGTEOIIR4DphtlpwAbEWzOhsC9wK33WDVaiRCiLpSyitSynXAuqrWpxoRZvQ/TUr5XAXUOQUzIwbtxlJadgPfArWBpsADgL8urzOaIXO3lPJEWRVVVF+klN+Ws/wB4EAFqVNZPGP0f0mVaWGbQuAg8AfaS8hdJRXQvbAYGyK7gdW68kN1aYPRDI03yltOxxI0IwYgWgjxprmBVKFIKdViZwFaAAVob8YSuAYMsCHbF+hullYLeBLN6MnQ1XUB+A14FnCzUo80Wqbr6t2i2/ZpNJedh052MNpbTg5wBvgK8Darr6dZnT2BkcBOXZ0XgWXAbVZ0GY/mMTkAnAPygatoF9NXQCsrZaabbc8DzTV7XLf/cTq5OCOZZLM6Guj28x8gW1fuLLAX7aE8xMp2vYCXgd+BS7oy54BfgNGAk5l8mJmeUWhdORuBK7rtJgCdy3DeOKyLbrvSzpLswPbM96XQ6H9fI7nWQLEVGQmE2akvzmx7dYCFZjK/lXAe9KzA63Kg7pxNAXJ1xysJmAs0u8muQRNdrOS3QvOUJOnO2RzgENqbewMr17NxfQPNjulho7ytgIsuL9EoPdHB89agrxXZMDO9BDAMzdtzGu0+kwlsQnsguljZ79t0x/uIbp/zdWV36dqjdynOpxBMrweL+5pOrgXwve54ZQPbdMff4vhW1Llu1D7uRutxRtuyeX/QnbN6uRSzOr4zyrsK1C1vOV2+E9p1oJcZXJFtYbGPlVn5zbCgufSNT86XSlHWH9hTwkX+JxBgVs44fw/XHzrGy0bgBRt1/mpWn/kFtsFGuUvAnWZld5egf675BYvlw2uT2XqcTs7qhahrt/QStptots3mwLESysQDtY3KhJnlb7HR1teAlqU47qXShcoxYpYb/V9rJBdrlL7MrEyYnfrirGzTBdhnJhdp5zzo6Wgb2tlPV7SHiL32GnSTXYPG+dPN8p4A8uzs20XgHrMyi43yTwP1dOlfmJVrbFQm0Sgv0cHz1qCvFVnjc80NLRbMXj0JQB2jMrcBWSWUsThn7ZxXY4zKXbAhc6fu+Fjb1g9m6+b3xJLayWq72dE3zkjW6v0BCDCr8xOz/KFm+Y+Wp5yZzEqj/JjyXvf2FtWdVDK9jf5LLF3z9lgItDdaXw9sBzoC9+vS7gIWAf1s1NEezQuyAhigKwta99W9wN/AKjT3fjtdXi8hRCcp5Q4bdfZGMywSdfU/oEv3Qbs4jHU+j/Z2dJTrHoX6wCNAI7Qb0KdAGxvbAuiG9sYZr5O/ZEcWtIskWPc/F5iP1g8fAIQCPYyFhRDOaG3Q1Cj5ezQvTm+gqy6tD9ob9ESscw/a2+wKtOOiP0a1gcnAUyXoXVZd9PECw4EOuvxLwLu6/2WJF/obzRvUF+gvhGiF5n14XJd/CO3BMaQMdQMgpSwUQszDtK++N9qxriz+x3V3NmgP2+/QHsbNgAfN5G+Ga9AqQohOwByuB6L+jebyF8BjaO3hC6wUQoRLKfXn0SS0eKZmaNfy50KI+WjeKj3jpZQnS1ChpPMWNE9kSXzI9WNRjGZc/412rY9Gu2f0BD7i+rU7DvDU/c9Eu0dkoMWBNQO6O7BdY4zld9mQWYB2fPT8hGbg9sTyvKsOmHc3HSthvS3avaqs5YzZCQzS/e9BZVKZFtLNsKC5DPUW5ZlSlLsDU2v1G7P8BWb57YzyjNMz0Lnr0FyZxnnnAC9dXiuzvOeM6utplhcPCDu6RJrp6q6r4wm0GIuX0Iw54zKNjOSnm+Utx6wrRycXZySTbJQ+xSh9nZVyTkBTo/UHzbb3tpnsr0Z5hYC/Li/MrNxJfXvq8v8wytvj4HEvky722sPB7Zrvy3Sgv9H6HGCG0fqT2H87Nq8vzsZ27zOT+8zOedCznNeiD1qXgb6+FCw9KJ5A4M10DZofV6N0Y0/aXqCWUZ4fWheLPn+K2b53wNSDY+zVmG2l7RON8hMduY7NZKyea2hGlnF3/ctm5Z4yyjO+dj8ySv/SyvZcgdBSnFvG+2fhOQA6mem/wChPYOlZ62lW/qVSLl1K0NeRNn/MTKfxZvnNzPK/LE85M5lRZsfN4v5fUYvyxFQeXc3W55utz0NzYeq5B82tbc4aKeUV3f9kK3lZuv9HzPJ87ei2UOrONB0LzHTpgO5tWjfW/02grp36QAtutjVq4V0pZXEJ5Y3ZjPZG5oTmRfgHbRTTUbQ3tF+llMeN5G22tZSyWAixgOujDpzRAlHXYMlCo/YELT5A/2Ztrz2NqShdyo2Ucr0Q4gBwO9obbbYu6wLaMR9eAZsxH2YprUo5UpEQXYAuVrL2Sy0I/G60h5OemVLK8yYbl/IqWj893CTXoB26Gf1vC+TZGS3bFe3BD4CUcrcQ4t9oXhC47tX4E9NRRJVNZ0wHmLwvhHjfhqzx9fIbmncUYJIQIhItTu8o8BfaPSLFai3WCTD6f8FKfkez9QX6P1JKKYT4GlOvvQlSyg9KoUtlYX5yODqFQ1nKGbehM5pRneHg9kqFMmJKJhXt7QsgUAjhJ6V0ZHign9n6mRLWzeX1pOn/SCnzzW5S6UZ5hWZ59uYAOlvCui+AEOIhjG58JWB1PhIdSQ7WAYCUco9uRNg7aG/frXSLniIhxP+klP/WrVdUWyebrecZ/Xd0TqWK0qWimIkW/OjO9eHbX0opc+w88EqD+Yi81HLU1Q+YZiV9AdooNvO2OlFCfTX+GiyB0pw7AVbS5qJ55zyN0r6UUuZZka0sSnv+BwBIKVcKId5GG2lXG+1lo52RXK4Q4iUp5WcO1lvSxWA+30lJx8+0cm24cmn4XUrpSFecPcyNMfMXUfN1vZFR1nLGmJ/7FXKzsYYyYkpmA9eNGIHmFp3pQDlzQ6c+pkMM65cgr8fe0LRCB/SwRlAJ65m638eM0rLRYhF+0z0A70frEy4RKWV2yVIWZT4XQsxF67tvjebC7IL2tuwMvCKEWCul3IT1tj5mtm6Mo21dFq9CRelSUSxCi1HQH+N8tBimciOEcEGLTTBmY0XUbQPztmpSSvmaeA3a4yIQqPv/J1rAri2sGZdfYGrAAPxXCPGzlLI8xmhpMG/zGDQPqC126/9IKV8XQryH5p1phXaP6IXmlXIHPtbdI45brcmU81x/UbJmWJkfjyBMzyXz42eO+XwrJTEDx+KJ7LHXbL252Xozs/W/ylnOGOM2LMK6d6tCUDP2lsxsTG9UbwkhrAYACiH6CCH0Lt6tZtnmN/vxZuvm8pXJaLPZFsea5esD2/yN0o5LKddJKfWzMT5GJSGECBZCBEsp86SUm6WUX0kpX0ZznxsHuepdvDbbWgjhhOn+FaEFdlYW1UkXdG/Vxm+jS6SU5h6IUiO0mannocWd6EmUUtoKiiwRKeV0KaWwskTpRLZhalC8IISoZ6ZXHSGE/sF+M1yD9thi9D8ELebnA+MF7YXrL8AkwFgI8QTa5Jyg3d/0XWF+wGJdgLqjGB+TOqUoB9r5b3x/dTPfB91+xKLNnfS3Tv8mQghfKWW2lHKjlPJTKeULaIHWepwx9c7Y46jR/0ZW8s2Ph+F46Y7jGKoZuq5WY0PoQSGEu9H6MKP/2WjTP5S5nBnGbXiilOEEpUJ5YkpASnlICPF/XB+BUQdYL4T4FdPJ7nqjudbHAZullPuEEL9wfcTDKCGEP9pF2wFtngs9G6WU1vriK4s+QKIQIgGI4PrICIC9Ukp9X/whtNEtAHcIIb5Fi03pienNoqK5B/hOCLFdt73TaDfKboC3kZz+Le4ntNE/rXXrrwohWmA5Igi0ANVK6ZuthrromY0WSwRlHzl0u84l7o72JmY82R1o7nRzI8Gcr4QQWVbS06WUD5WkgJQyUwjxOddjIUKBJCHE92jnSChaYPUkYNVNcg3a4wO0ESBOaJ6Av4UQy9AC1OsALdFGhgSgeShOAOhGqhlP3vk22kSGf6B1zXRD69Yzn8TMFsZemwAhRByal0Kixf7Y7GqRUl4SQsRwfeTfGJ1+G9CCjQPRDJEuaF13+knohqB5jTaj3adOo3nKB5htwlFP5yauG7URVvTcIYT4i+szXY/RGdC70drW7mgoKWW5u1OEEMZxNR2M/vua5X0hpdR7f6ehBZGDZlhsFkKsQhuBZDzKb5a8PnqtPOX0RBr9/83ObpWfyooYvtkWtJE5VzGNyra2RBmVCUBz89qT3wcEmW3L6miEsuZhOTLiRxu6XAbuMirXTJdmTdZ8dFJPo3LTjfPstGmckVyyUbr5PATWlsOYjiRqgXaTtlfmV0znmgizdezs6efAuVJqXcqzPRv7Mt2BMlFmZcLs1Gdv2YLZ5GXWzgM7S2nathalmyemRl+DDmyvpHliTK5PNCPUeG6f7Vyf0O4Zo/QiTK/pRKO8RDMd7sBy4kT90sGBc80dLVjX4fMEbQSPI+els4PnVQOzfQi3ItMO2/fDjdbauyIXB/bX6rYdaKsVgKuV7ZW1nPlkdxYTk1bkorqTHERKORftTe9faG8JZ9DiC/LQHlhL0R6+3xqVOY/WX/ssmjV6Ee1CuYR2gU1GG0ppNyisEvgQeBTNxZyD1t+7EugkpTT0h0rNmu8GrEVzG15Du+k9BHxdifr9DryCNufFYZ1+RWg3kD3AWzpdDW/1UsrDaG9J/9bt12W0ts5AO15RaDPXXqtEvaudLhVMMdr5op8VNQa4V0rZVd6g7yZJKfOllI+ieVyWo42Iy0M7P4+hnZd/G8nX6GuwJHT3pTvRvut2AK0ditD2cyfaXER90DwNoHUv6bsAs4HRUspCXV2foV3roD2IFuk8VyXp8DeaZ2QH2j2iVEgpc6WUD+jqWI0WSK2/t57U6fQvTL2/P6B5C9ahHfcrRvv9O9ps2X2llEUO6pCGaYzfMCsyf6INtV6Odrxy0Dwxo9DuSdUSqXXHdUEz/vVe7YtoL1KPoxkZFnFfZS2H5pXSxwidRTtWlYbQWU6KmxihfT04wSipl5QysWq0UShuPdQ1WP3RfQBSH9/xj5Ty9qrUp6YihPiK65MSvi2lrNQPQCpPjEKhUChueaSU8Vw3NFsLIczjaxQlIIQIQJuTCrQRSZU+P44yYhQKhUKh0HgRLY4D4D9VqUgNZTJacDjAm9J60G+FokYnKRQKhUKBIe5FvdyXESnla8BrN3KbKiZGoVAoFApFjeSW9MT4+/vLsLCwqlaj2pObm4u7u3vJggpFFaPOVUVNQZ2rJbNnz54MKaW1T2VYcEsaMWFhYezevbtkwVucpKQkWrZsWdVqKBQlos5VRU1BnaslI4Rw+OOdqu9PoVAoFApFjUQZMQqFQqFQKGokyohRKBQKhUJRI1FGjEKhUCgUihqJMmIUCoVCoVDUSJQRo1AoFAqFokaijBiFQqFQKBQ1EmXEKBQKhUKhqJEoI0ahUCgUCkWNRBkxCoVCoVAoaiTKiFEoFAqFQlEjUUaMQqFQKBSKGokyYhQKhUKhUNiluFiyJ+USB9IvV7UqJtySX7FWKBQKhUJRMkfPZTFvazLx/5zlfFYeD9wZzKcj21e1WgZuKU+MEOJBIcScy5erlyWpUCgUCkV15HJOIYt3nOR8Vh4AiYfOk1dYVMVaXeeWMmKklD9KKSd6e3tXtSoKhUKhUFQLzmXlcuFqntW8do188Pd0M6xfzSvk92MXbpRqJXJLGTEKhUKhUCgg9dI1vvrtGIM/30qndzeyaMdJq3JOToK+rYMA8KjlzMA7g/GtU+tGqmoXFROjUCgUCsUtxpYjGfx3bZJh/Zd/zvB873CrsmPuDqVf6yC6NK+Hm4vzjVLRIZQnRlEiQgiEEDg5OXHs2DGbcr169TLIxsXFmeRFRUUZ8oQQODs74+3tTbNmzRg0aBCffvopFy5Yd1EmJiYihKBnz54Vtk+nTp3ilVdeISIiAl9fX1xdXQkMDKRPnz58/PHH2IqbOnz4MM888wwtW7bE09MTDw8PbrvtNp5++mkOHTpktUxycrJhvz09PcnKyrIqJ6WkWbNmBtnExEST/J49e5q0oRACLy8vIiIiePfdd8nJySlXmygUipuLgqJiCoqKreb1bhWEENfX96ddIS3T+j2kVXBderUMrHYGDCgjRuEgLi4uSCmZO3eu1fwjR46QmJiIi4t9597DDz/MtGnTeP3115kwYQIRERHs2LGD5557jqZNm1oYP5VBbGws4eHhvP/++xQVFTFixAhefvllBg8ezJkzZ5gyZQrNmjWzKDd79mxat27NF198QXBwME8++SRPP/00DRo04Msvv+T2229n9uzZNrfr4uJCdnY2S5YssZq/ceNGjh8/XmIbjh07lmnTpvHGG28wbNgwjh8/zquvvkrPnj0pKCgoXWMoFIqbimv5haz9+zQvfLuXiLfi2XjwrFW5AC83Ihr7GtbvbOhNRpb1uJhqjZTyllsiIiKkomQOHjwopZQSkA0aNJAdOnSQ9evXlwUFBRayL7/8sgTkI488IgE5f/58k/yxY8daTZdSyoKCAvnVV19Jd3d3CcjFixeb5CckJEhA9ujRo9z79M0330hA+vr6yjVr1liV2bJli2zbtq1J2oIFCyQg/fz85G+//WZRZtOmTdLPz08C8uuvvzbJO3HihARkp06dZP369WWHDh2sbnfYsGHSzc1N3nfffRKQCQkJJvk9evSwmp6eni6DgoIkIOPi4kpogZsT/bmqUFR3KvtcffPHAzL0X2sMywtL/7Qp+/O+dLng9xMy7dK1StWptAC7pYPPc+WJUThMdHQ0Z86cYc2aNSbpBQUFxMXF0aVLF1q3bl3qel1cXJg4cSKff/45AFOnTq2UrpGsrCyef/55AJYuXcrAgQOtyt1zzz3s2LHDpNyUKVMAWLx4Md27d7co061bNxYtWgTAlClTrHYZubi4MG7cOHbv3s1ff/1lkpeRkcGqVasYMmQIfn5+pdqv4OBgBg8eDMDOnTtLVVahUNxc6INw9WxMOmezS+m+O4IZc3cYIT61b4RqlYIK7C0DYa/8VNUqlIrk96w/rEvLiBEjmDp1KrGxsQwaNMiQ/sMPP3Du3Dnef/99jh49Wub6x44dy4wZM0hJSeHXX3+1aWSUlWXLlnHx4kU6d+5Mv3797Mq6uV0fUrhs2TIuXbpEZGQk/fv3t1lmwIABdOzYkV27drFs2TLGjRtnITNhwgTee+89YmJi+PTTTw3pCxYsID8/n+joaGJjY0u9b9rLixa/pFAobj6klPxz+grrD5zllwNn+GxUe5oFeFrIdQj1xbeOK5euaV3LTgJSLmTTPNDrRqt8Q1CeGIXDeHl58dhjj7Fu3TpSU1MN6TExMdStW5dhw4aVq34nJye6desGVI5HYcuWLQD07t27TOX69OlTomzfvn0B2Lp1q9X8pk2bcu+997Jo0SITb5M+TqcswcunT59mxYoVAHTq1KnU5RUKRfXnhW/3MnD2FmZvPELSmSx+OWA91sXF2YlRnUJ5omsTvp3YmV2v9rlpDRhQRoyilERHR1NUVMS8efMASElJIT4+nlGjRlGnTp1y19+gQQMAzp8/X+66zDl9+jQADRs2LFO5Ro0alSirl0lPT7cpEx0dTWZmJt9//z0AmzdvJikpiQkTJjikT1xcHNOnT2fatGk88cQTtG7dmnPnzhEZGcljjz3mUB0KhaJmERHqa7K+/sAZm7Iv9b+N1x9oTaem9XBxvrkf86o7SVEqOnXqxB133MG8efN47bXXiI2Npbi4mOjo6AqpvzzdIpmZmXz00UcW6VOmTMHHx6fculUUjzzyCP7+/sTExDBmzBjmzJmDq6srUVFRDpVfsGCB4b+Hhwfh4eEMGTKEqVOn4urqWklaKxSKyuJSdj6/Jp3j10Pn+PDRtri7Wg5l7tu6Pq+vPmBY3592mUvZ+fh6VJ+J56oCZcSUgYqKMampREdH8/zzz7N27Vrmz59PREQE7dq1q5C69R6MgICAUpfNzMxkxowZFulRUVH4+PgQHBwMQFpaWqnqrV+/PqDNLVMSepmQkBCbMrVq1WLMmDHMnDmTbdu2sWzZMh566CECAwMd0ichIaFC58xRKBRVx5Slf/LjvtMUFWsvcIPbNaB3qyALufre7nRq4oeXuyv9bg+iT6ugW96AAdWdpCgDo0ePpnbt2jz55JOkpaUxceLECqm3uLiYTZs2AWWL7QgLC7M6BC8sLAyArl27Atp8LKVBX27Dhg0lyupl7rnnHrtyes/VsGHDyM3NrbA2VCgUNQsvd1eDAQPYjHUBWDqxM7FjOzCsQyP8lAEDKCNGUQZ8fHwYOnQoqampeHh4MGLEiAqpNy4ujpMnTxIcHEyvXr0qpE5jhg4dip+fH9u2bSvRIMnLuz7p09ChQ/Hx8WHnzp3Ex8fbLBMfH8/OnTvx8/Nj6NChdutv2bIl3bp1IzU1lbCwMENAsEKhuLm4fK2AhKRzNvP73W7qddlw8KyJUWOMGn1oiTJiFGXi7bffZuXKlaxfvx4vr/JFvhcWFhITE8MzzzyDEIJZs2bh7u5eQZpex8vLyzCj7vDhw1m/fr1Vue3bt3P33Xcb1uvWrcuHH34IwMiRI62OPPr9998ZOXIkALNmzXKoTebMmcPKlStZsWKFujkpFDcRUkp2JV9k6rd7iXx3AxMX7ibDxleiOzWph5e7Cy2CPHnu3ubEjYvESd0OHEbFxCjKROPGjWncuHGpy61atYrk5GQAsrOzOXnyJJs3b+b06dN4e3szZ84chg8fbrVsUlKSzeDXxo0b8+abb5a4/VGjRpGTk8Ozzz7LgAEDuOuuu+jSpQu+vr5cuHCBbdu28ddff+Hv729Sbvz48WRmZvLyyy/TrVs3evbsSUREBEII9uzZQ0JCAk5OTnz00UeMGTPGobZo2bIlLVu2dEhWoVDULP61fB/Hz2cb1pfvSWVSD8vPmdRycSLxpZ7U83SzyFOUjDJiFDeU1atXs3r1apycnPDw8CAgIIDIyEj69OnDyJEj7c5We/bsWZOROca0bdvWISMGtAnn+vfvz6effkp8fDyLFi0iOzsbHx8f2rRpw6xZsxg/frxFualTp3L//ffz8ccf8+uvv7J9+3ZAG7I9adIkJk+erIwShUKBEIIRHRvzzs8HDWlLd51iYvemVuWVAVN2hH5I661Ehw4d5O7du6tajWpPUlKSeigragTqXFXcaM5n5bFu/2ke7xxqtTv4wtU8Ov93IwVFkgY+tRnesRGTejTlxNEj6lwtASHEHillB0dklSdGoVAoFAoH2XIkg0U7Uoj/5yyFxZLwIC86N61nIVfP041X7mtF80BPujb3x1kFulQKyohRKBQKhcJBluw8ydr912fLXbrzpFUjBuCJrk1ulFpVRk5+EbVrWU7Od6NQo5MUCoVCoXCQxyJNPz/y8/4zZF7LryJtbjyXcwpYt/80r678mx7/S+CpRXuqVB/liVEoFAqFQkfqpWt8tzuV8feE4VPHckK5e5r508ivNqcu5tAswIMRkY1v6q6i/MJi/jx5iS1HM9h8JIN9qZkYT2Nz7koeeYVFuLlUjTdGGTEKhUKhuOXZdPg8sVtOsPnIeaQE3zqujLvHsjvIyUkw7YHb8a7jSodQ35tujicpJcfOX2XzkQy2HMlg+/ELZOcX2ZTPKSjiz5OZNrvUKhtlxCgUCoXilmdfaiabDp83rC/deYqoLmFWjZQ+rS2/bVSTybiax1adp2XLkQzOXMl1uKyrs+BERrYyYhQKhUKhqCoe7dCIWRuOGKb8P3Q2iz9PZdK+sW8Va1bx5BYUsSv5IpuPaIbLwdNXSlW+RZAnXZsH0K2FP52a+FGnVtWZEreUESOEeBB4sHnz5lWtikKhUChuIEfPXWXpzpM8fFcD7mjobZEfVNedXrcFsuHgWe5o4M1jkY0ID/SsAk0rnuJiyT+nr7DlqOZp2Zl8kfzCYofL+3u60S3cn67N/eka7k9Q3Yr/LExZuaWMGCnlj8CPHTp0iK5qXRQKhUJR+Ww7doFZ8YfZmXwRgKzcQt4feqdV2al9WzClTzhtGlgaOTWN9MwcthzJYPPRDH4/msGFbMdHULm7OtGpST3NcAn357Ygr2ob+3NLGTEKhUKhuLW4ll9oMGAAftyXzmsPtMLL3dVCtnVI3RupWoVyNa+Q7ccu6EYRneeY0XebSkIIaBPiTdd3RuVZAAAgAElEQVRwf7o19ycizLfKRhuVFmXEKBQKheKmpUeLAOrXdTcEq17LL+LHv04zslPpP2BbnSgsKuav1MtsOZLBlqPn+fNkJoXFjn9GqIFPbYOnpUszf/w8LIeT1wSUEaNQKBSKGsvfqZdZsuskHUJ9Gdy+oUW+i7MTwzo0ZPavR4ls4seIyEbc1ya4CjQtH1JKUi5cY/PRDDYfPs+24xfIyi10uLyXmwudm2ldRN3CAwirV6fadhGVBmXEKBQKhaLGsfdUJq+u/JsD6drImv1pl60aMQCj7w7jobsa0LyGBepmXstn69ELbDl6ns1HMki9lONwWWcnQbtGPloXUbg/bRv64OJ8803Sr4wYhUKhUNQ46nnU4h+jocH7Ui+zP+2y1aDcAC83ArzcbqR6ZSKvsIg9KZd0XUQZ/J12Gel4DxFN/T10XUQBdG7qZzXu52ZDGTEKhxk7dixr167lxIkTeHh4VLU6ZeKhhx7iwIEDHDx4kFq1amYfsEKhgEZ+degWHmAyQd3qvWk1amSRlJLDZ6+y+ch5thzNYMfxi+QU2J4d1xzfOq7c09zfYLg08KldidpWT24+35KiUti1axcLFy7klVdeMTFgkpOTEUKYLC4uLgQGBjJgwABWr15ttb7ExESDfJMmTZA2XjeuXr1K3bp1DbLJyckAhIWFWWzX3jJ9+nQA3nzzTU6cOMHs2bNLtf9xcXGGurp3725TLjk5GScnJ4OsOeZ6ubm5ERAQQPv27ZkwYQJr166lqMj6TSwqKgohBHFxcaXS3R7x8fGMGjWKJk2aUKdOHWrXrk3z5s0ZPXo0a9eutVpGSsmyZct4+OGHCQkJoVatWtSrV4+uXbsyc+ZMrl27ZrXc9OnTDfs9ZswYmzr99ttvBrmwsDCTPFvnW9euXRk4cKBNnRU1j2Pnr/Lid3/xzk//2JQZGdkIIaDnbQF8+XgELw9oeQM1LBsXrhWy4o9Upn67l07vbqT/R5t4+6eDJB46X6IBU8vFiXua1+NfA1qy5rmu7HmtL5+ObM/wjo1vSQMGlCdG4SCvvvoqdevW5amnnrKa7+3tzZQpUwDIy8vjwIEDrFmzhvXr1/O///2Pl156yWo5FxcXkpOTiY+Pp1+/fhb5S5cuJSsrCxcXFwoLrwexTZkyhczMTBPZuLg4UlJSGDt2rMXDr2fPngDcddddDBgwgHfeeYenn36aOnXqONoEBn03b97MoUOHuO222yzyY2NjkVJa6GvOtGnTACgqKiIzM5MDBw6wcOFC5s6dS4cOHVi0aBEtWrQolW6lISsrizFjxrBq1Src3d259957GTx4MK6urpw4cYKff/6Zb775hhdffJEPPvjAUC4zM5Nhw4YRHx+Pt7c3999/P2FhYVy8eJH169fz4osv8sknn7BmzRpuv/12q9t2cXFh2bJlzJ49Gx8fH4v8mJiYEtvP+HzLzc3l999/5+eff+bnn3/m448/5vnnny9nCymqirNXcnl/XRKr/kyjWGoBqVP73kbtWpZDfnu3CmLzy71o6Fu66/hGcz4rj+92n+LHv9JJOpMFpDhctmV9L0MwbscwP6vtcEsjpbzlloiICKkomYMHD0oppTx06JAUQsjo6GgLmRMnTkhAhoaGWuQtWbJEArJOnToyOzvbJC8hIUECcsCAAdLNzU0OHTrUqg6RkZEyODhYdurUSQLyxIkTNvXt0aOHBGRCQoLd/Vq6dKkEZExMjF05Y+bPny8BOWjQIAnIl156yUKmsLBQhoSEyI4dO8oGDRpI7fIyBbCaLqWUZ86ckY8++qgEZKNGjeTZs2dN8seOHSsBOX/+fIf1tkZRUZHs37+/BGSvXr1kWlqahUxubq786KOP5NNPP21Srk+fPhKQ/fv3lxkZGSZlCgoK5L///W8JyJCQEHnmzBmT/GnTppm04aeffmqx3YsXL0p3d3f5yCOPWD2vbJ1vBw8elPPmzZOA9PDwsDjfFDWH81m5suVra2Xov9YYlu92naxqtUpNcXGx3Hnignxu8R+y+X9+Mtkfe0vkO/Fy6rd75co/UuW5K7lVvRtVArBbOvg8V91J5SDslZ9MFlss3nHSRO7fK/bZlH3gk80msn+nXrYq93fqZRO5Bz7ZXO79scW8efOQUjJ8+PBSlRs+fDgeHh5cu3aNf/6x7hKuV68egwcPZvXq1Zw/f94kb9++fezcuZNx48bh4lJxTsOHH34Yd3d35s6dW+qyt99+O3fffTcLFiygoKDAJO+nn34iPT2d6OiyTQgdFBTE0qVL6dmzJ6dOneLdd98tUz0lsWTJEtavX0/z5s358ccfCQkJsZBxc3Nj8uTJzJw505C2ePFiNmzYQLNmzVixYgX16pl+8M3FxYV3332X4cOHk56ezmuvvWZ1+wMGDKBhw4bExsZa5C1cuJDc3NwytWFUVBQeHh5kZ2dz4MCBUpdXVA/8Pd0Y0yXUJC3h0Lkq0qb0XM0r5JvtKdz38WYe/XIbP/yVTkGR7ejcOrWcubdlIG880Jr4F7qz/d+9+XBYWwa1a1AjgpGrGmXEKEpkw4YNODs707lz5zLX4epqO0o+OjqagoICFixYYJIeExODEIInnniizNu1hru7OxEREezatYvLl60bifaIjo7m/PnzFvE+MTExeHp6MmLEiDLr5uTkZHj4L1myxGasUHmYM2cOAC+99FKJAdpubtdvojExMQC8+OKLdrvh3njjDeC6QWKOs7Mz48ePZ+/evezevdskLyYmhiZNmtCnTx/HdsYMfXvdDPNf3OxczbPdXTipezM8ajkTHujJpyPb8emI9jdQs7Jx+GwWb6zeT+d3N/Laqv26biNLnAS0beTDs72as3RiZ/a+0Y95UR0Z37UJ4dV4ev/qijJiFHbJzs5m7969tGrVqtQjkhYtWkR2djYBAQFW40f09OzZk+bNm5u8mefk5PDNN9/Qu3dvmjZtWmb9bdGxY0eKiorYunVrqcsOGzaMunXrGh7qAGlpaaxdu5bHHnsMT8/yzUXRtWtXXFxcOHfunCGQuaIoLCxk+/btAPTu3btM5UoyMFq3bk1ISAh5eXkWRoqe8ePH4+TkZNKG27dvZ//+/TzxxBNlupHPnz+fa9eu4eHhYTMeR1H1nL2SyzTdwz490/q8J34etVj2VBfWTenOA3eG4ORUPR/sBUXF/LTvNI/N2Ua/WZv4eluKTeMsrF4dXhvYiqXDw1j9zD281P82OjetRy0X9RguDyqwV2GXtLQ0ioqKCA62P8NlZmamYQRQXl4e+/fv56effqJWrVp89dVXuLvb/uqpEIIJEybwyiuvsGnTJrp3786yZcvIzMwsc9dMSdSvXx+AkydPlrqsh4cHI0eO5KuvviI5OZmwsDDmzZtHUVFRhejr5uZGvXr1OHv2LOfPn6dJkyblrlPPxYsXyc/XPgTXsKH1icFKKteoUaMS5Rs1akR6ejrp6elW80NDQ+nXrx9Llixh5syZeHh4EBMTg7OzM+PGjSuxfuPzTR/Yu3mz1qX67rvvUrv2rTlSo7rzReIxZm04bPiC8mcJR3nnkTusyrYKrr7fMTp9OYclO0+xZOdJzmfl2ZRzEtCnVRCj7w7lnmb+ODkJkpKSbqCmNz/KiCkHye8NdEhuZKfGDn+nY81z3RySu6Oht8PbLw8XLlwAwNfX167c5cuXmTFjhkmam5sbq1evpn///iVuJyoqitdff52YmBi6d+/OnDlz8Pf3Z9CgQWVX3g5+fn4AZGRklKl8dHQ0X375JXPnzmXGjBnMnTuXO++8k8jIyArRrzzdIomJiSQmJpqkhYWFERUVVQGaVRzR0dGsW7eOpUuX8uijj/Ltt98ycOBAQkJC7I5MAtPzzdnZGW9vb+677z6effZZ7r///huhvqIM+NZxNRgwAN/tPsVTPZtV+9FFoF2Tvx+7wMJtKcQfPEuRne8U+Xu6MSKyESMiGxNyiw59vlEoI0ZhF/0brbXYBmNCQ0MNXR9XrlwhPj6eCRMmMGzYMLZt20br1q3tlg8KCuLBBx9k+fLlPP3002zZsoUXX3yx0iaky8nR3NhlfWNv37497du3Z/78+XTu3JmUlBQ++eSTCtEtNzeXixe1r+4GBASUunxiYqKFQdmjRw+ioqLw8/OjVq1a5Ofnk5aWRrNmzRyq07jcqVOnCA8Ptyt/6tQpAKtBw3oefPBBgoKCiI2NpaCggOzsbIc9WcbnG0BSUhItW1b/OUJudYZENOSzxKOcuqhdf4Fe7qRn5lZrI+ZyTgHL96TyzY4UjpfwZejIJn6M7hxK/9vrq26iG4RqZYVdAgMDgeseGUeoW7cuQ4YM4ZtvvuHKlSuMGTPGoQDViRMnkpOTw7BhwwAqrSsJru+Pfv/KwsSJE0lLS+PJJ5+kdu3aPP744xWi25YtWygsLCQoKMhivhtHmD59usUwRL1nxsXFxRCgvXHjRofrdHFxoVOnToAW6G2PgwcPkp6ejpubGx06dLAp5+rqyrhx49i+fTvvvPMODRs25L777nNYJ0X1JPNaPikXrD/sXZ2deP7ecIK93Xl7UBsSXupJZBO/G6yhY+xPu8wry/fR6d0NvLnmH5sGjEctZ0Z3DmX9lO58N+luHmwbogyYG4hqaYVdgoODCQgI4NChQ6UuO3DgQAYMGMCePXtYvHhxifJ9+/YlNDSU1NRUunfvbjcYuLzo+6XvuuuuMtcxcuRIPDw8SE1N5dFHH7U6cVtpKS4u5p133jHUXxlMnDgRgA8++MDm7Lp68vKu9/dPmDABgJkzZxo8WdZ4++23ARg9erTdWCh9nUIIUlNTGT9+PM7OaiKvmsrlnAJmxh+m2/sJ/Gfl3zblHmnXgISXevJ459Bq97DPLShixR+pPPL5Vh74ZAtLd50it6DYqmyLIE/eevh2tv+nN28NasNt9b1usLYKUN1JihLQT7O/fPlyjh49SvPmzUtV/q233mLdunVMmzaN4cOH253vxcnJiRUrVnDy5ElatWpVXtXtsn37dvz9/WnTpk2Z6/Dy8mLdunVkZGQQERFRbp3OnTvHs88+S2JiIo0bN+Y///lPueu0xogRI1i4cCHr16/n4Ycf5uuvv7YI3M7PzycmJoaDBw/y6aefAjBq1Cji4uJISEhg6NChfPPNNyaxUkVFRcyYMYPFixcTHBzMW2+9VaIuzZo1Y926dVy7do1u3RyLB1NUP85czqXfrN+4kqvFMm09eoEdxy/QqWk9C1kXZydcqpmtevLCNRbtTOG7Xae4dK3AppyLk2BAm/qM7hxKZBM/NRy6GqCMGEWJDBkyhOXLlxsmSCsNHTp04OGHH2b16tXMnTuXSZMm2ZXXx5pUJocOHeLkyZNMnDix3Dehrl27lqmcfmRNcXGx4bMDW7ZsIT8/n8jISBYtWoS/v7/VsrGxsRaBu3pGjhxp9fMNxjg5OfH9998zevRoVq9eTdOmTenduzetWrXC2dmZ5ORkfv31V86fP2/yuQhnZ2eWL1/O0KFD+fnnn2natCkDBw4kNDTU8NmBEydOEBYWxo8//mgYAVYSJemrqP7U93anTQNvfj92vdt51obDLJ14dxVqZZ+iYslvh8+xcFsKiYfP2/1adLC3OyMjGzM8shGBXva9i4obizJiFCUyZMgQAgMD+frrr3nmmWdKXX7GjBn88MMPvPXWW4wdO7bELobKRj+pnq3vQN0I9IG3tWrVwsvLi9DQUMaMGcOQIUPo168fTk623exbt261Ob/NXXfd5ZBR4OXlxapVq/jll1+Ii4tj27ZtbNy4ESklISEh9OnThzFjxjBgwACTcr6+vmzYsIHvv/+ehQsXsnHjRi5cuICnpyetWrXimWee4amnnir1N6kUNZ8X+rbg92PbAG0UUo8WgRQVS5yr2RwvF67m8d3uVBbtSCH1ku1uUYBu4f483jmU3i0DcXGuXl1fCg1RGTOCVnc6dOggbU3CpbiO8YiP//73v/znP//hjz/+oF27dlWsWdnJy8ujadOmtGrVqsQAVUXNQY1Oqnxy8otY/kcqwzs2wtXGA/25JX/Ssr4XY7uE4elWfd6RpZT8cTKTb7an8NO+0+QXWY9zAajr7sKjHRoxqlNjmgaUb+JKa6hztWSEEHuklLZHBRhRfc4yRbXmhRde4Msvv+SNN97gxx9/rGp1yswXX3zBmTNn+Pnnn6taFYWiRpBbUMSSnSf5PPEY57PyqOXsxLCO1ic8/GRE9XrBuZZfyOq96SzclsI/p6/YlW3ToC5jOofxYNsQ9aXoGoQyYhQO4e7uzsKFC0lISCA7O7vUnyCoLri5uTF37lzatm1b1aooFDWCmfGHmbPpuGF99q9HGNSuQbUbWWTM0XNX+WZ7Csv/SCUr1/bEibVcnHjwzhBG3x1K24beKlC3BqKMGIXDdO/ene7du1e1GuWiKuNgFIqayNguYcRtTTZ0waReymHZnlSHZyG/URQWFRP/z1kWbk8xCTC2RmO/OjzeuTGPRjTC16NyJtRU3BiUEaNQKBQKmzTwqc1jkY34elsKtV2dGdsljAFtHBt5diM4eyWXpTtPsXhnCmev2P6OkRDQu2Ugj3cOpXt4QLX9qKSidCgjRqFQKG5hCouKWflnGlm5hYzvav1jo0/3bI67qzMTuzfF39PtBmtoiZSS7ccv8s32FNYfOEOhne8Y1fOoxfCO2neMGvnd2FFzUkrOXMnlyNmrHDl3leSMbM5duIj3/jz+31DrXdoLtyWz5WgGUkKx1OoolpJiCcW6gTjFUlJcrP1KXboE/jf0TqvByD/+lc4nvx4x1KEvo69H6srrtyONtvfGA60Z3N7yY7F7T2US6OVW5d+GUkaMQqFQ3IIUFUt++CuNjzccIfnCNerUcuahu0KsGin1vd35z/2VOwGlI1zJLWDlH2ks3J7C0XNX7cp2CPVl9N2hDGhTH7dKnl2vuFiSlpnD0XNXOXIuy2C0HD13lat51mJysmwaMfvTrrD+wNky6XEtv8hq+uWcAg6ftd9etsgrtD6SS3DdqKpKlBGjUCgUtyD5hcW881MSGVe1Lphr+UXM2XS8Whgr5vyTfoVvdqSw6s80mw9qgDq1nBnUrgGPdwqldUjdStfrtVV/89epyxw9d5WcAtt6WUNKaTWQ2M4UUSViy6hwKkfAsr06q4ENo4wYhUKhuBWpXcuZp3o24601/xjSluw8yZQ+4dSpVfWPhsKiYn76+zQLt6WwO+WSXdnmgZ6M7hzKI+0bUNfdtdzbLigqJuXCNY6eyyKnoIhH2ll2pwD8nXaFv9Mul2kbUmpxOuaUZ4SULaOiPOE/tnrqhFCeGIVCoVBUMrbe+AFGdWrMV78d40J2Po9GNOSZXs2r3IApKCpm1Z9pfJZwlOQLtj9Q6uwk6H97EI93DuXupvXK9PDPKywiOeOaoQtI3x10IiObgiLtAV2/rrtNIyY80JO/TmXa3YanmwvNAz0JD/SkeaAn1y5foIHZt8qMGRnZmG7N/RFC4CQ0j4eTk2bcCHTrujwhBEIvI6BZoPXJ+Qa0qU+7xr6GMoZ6deWFsKxXL2NrzpzWwXWtGmE3GmXEKBQKxU2IlJKNB8/x0cbDvD/kTm4P8baQcXd1Zuawu2jkV5vQelU791N+YTEr/0zls4RjnLxo23gJquvGiMjGjIhsTFBdxz9hcuhMFklnrujiVbI4cu4qKReuUWQnKBjgzJVcruQWWPXwhBsZDd61XQkP9CQ8yJPmgV6G//XrupsYWElJBbRsaX2yQIA2Dbxp08DyWJUHnzq18KlTsUPJq8voLmXEKBQKxU3GruSLvLXmH/alal0dH284wpwx1mdx7xpu/UOjN4q8wiKW7Unl84RjpGXa/pZRl2b1GN05lD6tg2x+9iC/sNjmJHyvrNjHnyfte01scfTcVdo39rVIv/+OYO5o4E3zIE8CPN3UZHlVgDJiFAqF4ibjSk6BwYAB+OWfs/ydepk7GlbsG355yC0o4vvdp/gi8Rjpl3Otyrg4CQa3b8DE7k1pHuhlSL+SW8DRc1c5auRVOXL2KgVFxex8tY/VusIDPUtlxAR6uREe5El4oJfNOJtGfnVu+LBthSnKiFEoFIqbjHtbBtK2kY9JvMbvxzKqhRGj/xbTl78dszk5nauzYGhEI0Z3bkx2fhE7T1xi0Y6TWszK2aucuWLd6AHIvJZvtesk3MgIMibE253mQbruH313UIAX3nXKHyCsqHyUEaMoEXMXqZOTE76+vtx5551MmDCBkSNHVpFmpUe/L0IIjhw5QrNmzazK9erVi8TERADmz59PVFSUIS8qKooFCxYY1p2cnPD09MTf35877riDPn36MGLECOrVq2dRb2JiIr169aJHjx6G+svLqVOn+Oyzz4iPj+f48eNcvXrVcHwefPBBoqKi8Pa2fHgdPnyYjz/+mI0bN5KamoqUkoYNG9K7d28mT57MbbfdZlEmOTmZJk20CdE8PDw4ffo0Xl6WDwcpJc2bN+f4ce2bOwkJCfTs2dOQ37NnT3777TeTMp6enrRo0YIhQ4bwwgsvULt21U6iVZMRQjC1bwvGztvJgNvrM7lPOK2CK3/IsT1y8otYtCOFrzYd53yWdeNF+7hkQ4olbD2awZKdJ0u9naPnrtIhzM8i/c6G3tzbMtAQYBse5EWzAA+8KmA0k6LqqNFGjBDiaeD/gGDgADBFSrm5arW6eZk2bRoABQUFJCUlsXr1ahISEti9ezczZ86sYu0cx8XFhcLCQubOncu7775rkX/kyBESExMNcrZ4+OGHueuuuwDIysri1KlTbN68mdWrV/Pqq6/y8ccfmxg/lUFsbCzPPvsseXl5tG3blhEjRuDr68uFCxfYsmULU6ZM4a233iIjI8Ok3OzZs5k6dSrFxcX06NGDBx54ACEEe/bs4csvv2TOnDnMnDmT559/3up2XVxcyM7OZsmSJUycONEif+PGjRw/frzENhw7dixhYWFIKUlNTWXFihW8+uqrrF69mi1btuDqqh4w1pBS8sNf6SQeOs/MYW2txmJ0D/dnw9TuJt0wVUF2XiHfbE8hZvNxMq7mW5Wp5eLEyMjGTOrRlGDv2kz9di8pdkYm2cJJQFpmDtaifzo1rUenppYvFoqaTY01YoQQw4GPgaeBLbrftUKI1lLK0pvvihKZPn26yfrGjRvp27cvH330Ec8//zxhYWFVoldpCQoKIjg4mPnz5/Pmm2/i4mJ6GcTGxgLw4IMPsnLlSpv1DBo0yMJIKSwsZN68eUyePJlx48bh5ubGiBEjKnwfABYtWkR0dDS+vr4sX76cgQMHWshs3bqVZ555xiTt66+/ZvLkyfj5+bFy5UqLj3pu3ryZQYMGMXnyZHx9fRk9erRFvREREaSkpBATE2PViImJicHNzY17772XtWvX2tyHqKgoEw/N22+/Tbt27di5cyeLFy9m7NixJTXDLcffqZeZ/uMB9ujmTunXOoj77rAcsiuEqFID5mpeIQt+P8GcTSe4nFNgVcbNRfB45zAmdW9KoNFIo9sbeLPizzSbdbs4CcL8PQxdQPruoCb+Hri7Vu7svIrqRfX9lnrJTAXipJQxUsqDUsrngNOA+kzxDaJ37960bNkSKSW7du0CtO4GIQRRUVEcPnyY4cOHExgYiJOTk6H7ZM+ePUyePJm2bdvi5+eHu7s74eHhvPjii1y6ZHtSq2+//ZbevXsbyoSFhTFixAh2795dat2jo6M5c+YMa9asMUkvKCggLi6OLl260Lp161LX6+LiwsSJE/n8888BmDp1Kjk5tkdclJWsrCyDl2Tp0qVWDRiAe+65hx07dpiUmzJlCgCLFy+2+lXybt26sWjRIgCmTJlCVlaWhYyLiwvjxo1j9+7d/PXXXyZ5GRkZrFq1iiFDhuDnZ+nWt0dwcDCDBw8GYOfOnaUqe6vw3rqDBgMG4J2fD5JbytliK4PiYklyRjbf7z7FsC9/564Zv/C/9YdtGjAAn4xoz+sPtDYxYADu0A0xdhLQIsiTB+4M5oU+Lfh8VHviX+jOP28OYMPUHnzxeART+93GQ21DaBVcVxkwtyCV5okRQgwFegB3AW0BL2CRlPJxO2UaAm8CA4B6aEbJKmCGlPKSkVwtIAL4wKyKX4AuFbgbdgl75acylWvToC5rnutmNe+BTzazP+1KmepNfs/6g6wykboZG83d2ceOHaNTp060aNGCUaNGkZOTQ926Wp98TEwMK1eupEePHvTp04fi4mL27NnDzJkzWbt2LTt27DCJs5BSMm7cOBYsWIC/vz+DBw8mICCA1NRUEhISuO222+jQwfrwUVuMGDGCqVOnEhsby6BBgwzpP/zwA+fOneP999/n6NGjZW0Wxo4dy4wZM0hJSeHXX3+1aWSUlWXLlnHx4kU6d+5Mv3797Mq6uV3/Fs6yZcu4dOkSkZGR9O/f32aZAQMG0LFjR3bt2sWyZcsYN26chcyECRN47733iImJ4dNPPzWkL1iwgPz8fKKjow1erdJg65xSaEx78Hbu+3izYX6Ts1dy2ZNyiXuaV81Q6VV/prFk50kOpF+x8Z0g26Resm7g39nQmxVPd6FV/bo2J1tTKKByPTGvAc+iGTG2/YI6hBDNgD3AOGAnMAs4DkwGtgkhjDsz/QFnwPwrWWeB6vON+JucDRs2cOjQIYQQdOzY0SRvy5YtPPXUU+zYsYNZs2bx5Zdf0r59ewD+/e9/k56eznfffceHH37IrFmz2LRpE7GxsRw8eNDgxdATExPDggUL6NixI0ePHiU2Npb//ve/LFy4kJSUFCZNmlRq3b28vHjsscdYt24dqampJtuqW7cuw4YNK0OLXMfJyYlu3TRDtTI8Clu2bAE0b1hZyvXpY30YqjF9+/YFtC4pazRt2pR7772XRYsWmXibYmNjCQ8PN+kmcpTTp0+zYsUKADp16lTq8rcCLYK8iOoSBkCfVkH88kKPSjVgCoqK+Sf9ik1vz3/cDdoAACAASURBVMmL2ew4cbFUBkwtFyfaNvKhbm3rMU/urs60b+yrDBhFiVRmTMwLQCpwFM0jk1CC/OdAIPC8lPITfaIQYqaurneAJytHVYUj6GNiCgoKOHToEKtWrUJKyQsvvEBoaKiJbFBQkCEQ2BxzWT3jx49n6tSprF+/nn/961+G9E8+0U6Hr776ymKUjbOzM8F2pvC2R3R0NHPnzmXevHm88cYbpKSkEB8fz6RJk6hTp/xzPzRo0ACA8+fPl7suc06fPg1Aw4bWp0MvqVyjRrZnDNWjl0lPT7cpEx0dzcaNG/n+++8ZM2YMmzdvJikpiffff98hfeLi4khMTDQJ7M3MzCQyMpLHHnvMoTpuRi5czeNERrbVUTYAU/qE06NFAN1bBFTodnMLijh8Nou/0y6zP+0KB9Ivk3Q6i/yiYpY/1YWI0OsTvl3Mzid283HmbTlht05XZ0GbEG/aNvLRzUZbl2YBnjYnrFMoSkOlGTFSSoPRUpJbWOeF6QckA5+ZZU8DJgKjhRAvSimzgQygCAgykw0CzpRLcYVNZsyYAWjH08fHh27duvHEE0/w+OOWPYRt27Y16cYwpqCggK+++oqlS5fyzz//cPnyZYqLr3/uPS3tuuMuOzub/fv3ExQURLt27ezql5ycTFxcnEW6eUCynk6dOnHHHXcwb948XnvtNWJjYykuLiY6OtrudhylPN0imZmZfPTRRxbpU6ZMwcfHp9y6VRSPPPII/v7+xMTEMGbMGObMmYOrq6vDo7KMh6p7eHgQHh7OkCFDmDp16i05Mqm4WLJ01yneX5eEq7MTCS/1sDoE2MvdtdwGzLX8Qg6evsL+tCvsT7vM/vQrHDmbRaGNafj3p10mItSXjKt5xGw+zsJtKTa/KF3LxYk+LQOZ1KMZbRp441xNpqhX3HxUl9FJvXS/v0gpi40zpJRZQoitaEZOZ2CjlDJfCLEH6At8byTeF1h+IxSGyolBsRUrUx2Qpfhiaf36tnv1hg8fzsqVK2natCkPP/ww9evXNxg8H330EXl51+eQyMzUJuvSezXskZycbDC0jLFlxIDmSXj++edZu3Yt8+fPJyIiokRjyVH0HoyAgNI/bDIzM63uS1RUFD4+Pgbvk7HB5wj643Lq1KkSZfUyISEhNmVq1arFmDFjmDlzJtu2bWPZsmU89NBDBAYGOqSP+fwxtzJFxZIRc7azM/miIe2jDUd4/YHSB5jb4q9TmcT9nsz+tMscO3/V5heKrbE7+SKnLl7jmx0p5BYUW5Xxqe1CdPdmjLk7VM2/orghVBcjRj+r1mEb+UfQjJgWwEZd2kxgoRBiJ7AVraspBPjSWgVCiIloHh1CQkJISkqqGM1vYjIyMkzayZE20z9Ur1y5YlV+//79rFy5krvvvps5c+aYDG8uLi7m/fffN8xDA3DtmjZXxMmTJ0vcfv369Tl48KBFunm5wsJCQ1qnTp1wd3dnwoQJnDlzhokTJxry9HOrnD592qSOy5cvW003pri4mI0btVPV+Hw7efKkYb9K2h9r+5Kbm0tSUpJhkr41a9YwatQou/UYY1zO2tBpY/Qjt5o2bWrQVX98c3JyDGm9e/dm5syZDB48mNzcXO677z5D3pUrWpC6+fErzXF1BPNztabSxKsY4wiq+VtPEOlfSKiP4x/vy8orwkmAh5V4kgPp11j55+lS6+XuIvjp79M2jR5vdyeG3u7DwJbe1HEtJC35WKm3catws5yr1YXqYsToAx0u28jXpxv86FLKb3XBvq+hTXa3H7hfSplirQIp5RxgDkCHDh1ky5YtK0Lvm5qkpCSM28mRNnN314ZKent7W5Xfu3cvoI0OatOmjUne9u3byc3NxdXV1aRsmzZt2L9/Pzk5ORXiJXFxcTGp/9FHH2XhwoV4eHjwwgsvGEZG+ftrwZLBwcEm8vq4HPN0Y+bNm8fp06cJDg5mzJgxhnY5c0br7axTp45D7WmL5557jg8//JC9e/eSmppqN1A3Ly/P4Ol67rnn+OCDD9i3bx+nTp0yBO+aEx8fz759+/Dz8+O5554ztIl+P2rXrm3Qv2XLlnTr1o3NmzcTFhbG+PHjDV1o+hFpjRs3NtlffcyReXpZMT9XayrTmxaxNfU3Ui/lUNvVmcl9wunVoYnNjxoaU1hUzHtrk4j7PYXpD93O43daxp4FNsrnP7/YN2Ia+tamTYg3ofXqcPTcVTYdOU9uoXXrxd/TjUndmzKqc2Pq1Kouj5Pqzc1yrlYXanRklZTycyllmJTSTUoZIaXcVNU6KeyjnxDPfMr9c+fOWUzKpkc/H8qkSZMMXhA9xcXFhmDVsvL222+zcuVK1q9fb3UK/dJQWFhITEwMzzzzDEIIZs2aZXjwVyReXl7Mnj0b0Lrn1q9fb1Vu+/bt3H333Yb1unXr8uGHHwIwcuRIqyOPfv/9d8OnJGbNmuVQm8yZM4eVK1eyYsX/Z+++w5uuFjeAvydJ0z3ohBZKgbYUEGhpWQVEhgsZDhRFGSKgAoL7er3ue/1d51UUHIgoIMOFIE5UUNlQoOyyKaNQ6N5txvn90UHTJLQZbZPm/TxPnzTfcXKqX8rLmas4NdoGnmolnrwxFm5KgeFdQ5HUvhXclPX/96zQ6jF75R4s3HQKWr3EwQzT/x4M9FYjIuDKdg4dgr0xskcbPHNzHL54oC9SX7geK6f3Q5CPGp9tPo0/0i5BozMOMKG+7nhhZFdsfHoIpl3bkQGGmo2jPHnVf+LM7U5Wfdy6fdTJYfTu3RsDBgzAqlWrkJycjIEDByIzMxM///wzOnfubHL8xdSpU7Fx40YsXboUMTExGDNmDEJCQpCRkYH169djypQpVx33Up/IyEhERkZafN/q1atx+vRpAJUDkM+cOYONGzfiwoUL8Pf3x4IFCzBu3DiT96alpZkd/BoZGYlXXnml3s+vXn9n1qxZuOmmmxAfH4/k5OSabQe2bt2KvXv31rQoVZsyZQry8vLw9NNPY9CgQbjuuuuQmJhYs+3Ahg0boFAo8O6772LixIkN+m8RFxfHf102gE4vsXjLaWQXl+OpG03/98op0kCjk1i79wLW7r2AoXGhWDS5t8lrgcoZRQ99sQt/HrkyC+5qa029NLob/DxU6BruZzBu5WxOCV77OQ3f7DpndnBvG38PPHxdJ9yV1I4Ly5FDcJQQc6TqNdbM+ZiqV3NjZshJKJVKfP/993juuefw008/4b333kNERASmTp2K5557zuQquUIILFmyBDfeeCMWLFiAr776CuXl5WjTpg0GDRqE0aNHN8NPAqxZswZr1qyBQqGAt7c3QkJC0KdPHwwfPhzjx4+/6mq1mZmZBjNzauvZs2eDQgxQGfBuvPFGzJs3D7/99huWLVuG4uJiBAQE4JprrsE777yDKVOmGN33+OOPY8SIEZg7dy7Wr1+Pbdu2Aaicsv3ggw9izpw5DCV2tv9cPp79bj/2n8+HQgA3dmuNHm2NZ5p9veucwfveZqZZA5VL+09dvBPbTuYYHD92qRA6vTQ5K+j6roaTOk9nFWP+huNYted8zQJ6dUUEeOLh6zrhzqS2cFcxvJDjEJbMOLH6Q4S4DpXrxJhcsbdqivVxVE6x7lR7hpIQwheVK/cKAKFVU6xtkpSUJK1Zqt7VsO+WnIWjP6s6vcSwt//E6VqbGl4T4YfVMwZAVWu9lIMZ+bjlvU017xUC2PrPYQirWpb/UmEZ3vntKJ4f2RUVWj0mfbYTe88aNlB3CPbGF1P7GnQbmXLychHmbTiONakZZsNL21aemDUkGrf3atugcTlUP0d/Vh2BEGKXlLJBy7A7REuMlPKEEGIdKmcgzQTwfq3TLwPwBvCxPQIMEVFTUyoEXhzdDfd/trPmWHpWCY5mFqFruF/NsW/qtMJcGxtSE2DSLhbggc9TcD6vFJcLy3EutxRpFw33tYpr7YslD/RBqK/5cVjHLxXi/fXHsXZvhtnZRu2DvDBzSDRuS4jgonTk0Bpz76RbAVRvSlO9aEh/IcTnVd9nSSmfrHXLDABbALwnhBgG4DCAvqhcQ+YogH/ZoU6jAIyKjo62tSgiIosM6RyKkT3a4Id9F3BLjzZ4sc7GhxVaPdakGq6OPDaxckXmPWdyMeHTHTVL+/9++JJR+T3bBWDx/b0R4GV6OvaRi4V4f/0x/Lj/Asw1wHcM9sasodEY3TPcoIWIyFE1ZktMPIBJdY51rPoCgHQANSGmqjUmCVc2gByBym6kuaizAaS1pJRrAaxNSkqyz5KsRER1FJVr4eN+5VerlLJmxtYLI7vijl5tMSTOeDHA9WmXkFNcUfPe39MNw7tUjl+JDvVB21aeRi0v1fp2CMSnk3sbfG61QxkFeH/9Mfx8wPxi5p1CvDF7WAxG9gjn6rrkVBpz24GXALxk4T1nUbkBJBGRU7lUWIYX1xzEhfwyfPtwMhSiMpi8+/sxLJyUhDA/D4RWfZmycucZg/eje4bXzADy9XDDp5N7Y8y8zcgqKje47rrOIfjw3kSjzRIPnM/He38cw7pDdffJvSI2zAezh8Xg5mvaMLyQU3KIMTFERM5KSolvdp3Df348jPxSDQDg9V/ScCijAJuOV676/NavR/DmnT1N3r9o0yl4uysNpkgDV7qSqkUEeOLTSUkYt2BrzbL/N1/TGnPvTjAYdLv3bB7e++MY/kgz7nKqFtfaF3OGxeDGbq2hYHghJ8YQQ0Rko98PZ9YEGABY8PdJg/Pf7D6HSclRuCbCcCms3OIK/Pfnw0YLysW3C0CPtsbLZvVsF4C1swbim13n0CnEB3cktoVSIXCpoAw/7LuANXszjGYr1dYt3A+zh8Xg+i5hDC/UIjDEEBHZQAiB527pgg1HLqNCa3pjRADYfirHKMT8sP+CyRVxH7y2o9mVj2PCfPHPEV2QX6rBt7vP4fvUDGw5kXXVzRx7tvXH7GExGBoXyhWVqUVhiCEispKUEj8fuIj//nzYbIBJ7hSE527pajCVutp3u88ZHWsf5IUebQNw78Jt2H8uHx5uSvi4q+DtroKXuvJ7rV5i64lsVOjMhyYASIgMwJxhMRgcG8LwQi0SQwwRUQNIKXEmpwQVWj1OXC7Cxfwy/LT/InaczjF5fYdgbzw7oguGdzHf+vHkDZ2xas95/Lz/AoorKnef/teILpixfHdNt1BBmRaXCstN3m9O/45BmDGkEwZGBzO8UIvmUiGG68QQkSmFZRpkFpThYn45LhaUIa+kAlMHdaw5fyG/FP/4dj/2ns1D76hWJtdpqebnocKc4bGY0K99vavcJkcHIzk6GK+M6YYD5wsQHuCB73afv+q4FnO6hfthTHw4RvYIR3g9q/UStRQuFWK4TgwRVdPq9Fi7LwML/j6FwxcMN0wUApiUHAU3pQKr95zH82sOoLCscqG5I2bWalEqBCb0a485w2LQytv0gnPmeKlV6NMhEAfO52PuH8cafF/7IC+M6RmO0fHhiA61bQd2ImfkUiGGiKhcq8M3u87ho79O4GxOqclrpAQuFZYjIsATlwrLagIMAJzNNb5naFwonh3RBdGhPlbXq0yjw2NfphrsIB3krcaK6f2gEEBRuQ7F5VoUlWtRptGhY7APronwY3cRuTSGGCJyCSUVWizffgafbDyJzIL6x5hczC9DRIAnHhjYEesOZiIlvXLR8M5hvmjl7YYOwd5o7eeJgTFBSGxvfqfphvrwzxM4dqnI4Nh/b++O2DC2sBCZwxBDRA6ttEIHvZTwNrGkfkPkl2qwZMtpLNp8CrklGpPXKAQQHuCJNv4eCPPzQGs/DwRWdQkpFQJv3dkTt3+4BbOGRGNycpTd11jR6SVW7DBcsXdsYlvc0K21mTuICGCIISIHdamwDPPWH8fKnWeh0enRs20AhsaFYkjnUHQL92tQkNhw5BIeWb6nZuPEutyUAmMT2+KhwZ1QWKY1WselWlSwNzb9Ywi81I3zK3Pn6RyDGUjeaiVeGNW1UT6LqCVhiCEih5JfqsGCv09g0abTKNXoao6nns1D6tk8/O+3owj1dceI7m0wJj4c8e0CzJbVrY2fybVUPNwUGN+nPaZd2wFebiq8vPYgVu05j0WTkzA0LsxkWbYEmHKtDgfOF6Bce+XnEbgSwr7Ynm5w/fVdw+Dn4Wb15xG5CoYYInIYBzPyMf6T7QZL+JtyqbAcn285jc+3nEbbVp4I8wReaxWBmDrjR0L9PDAuqR2WbqsMCb7uKkxKjsL9A6IQ5OOObSez8ejKVFwsKAMAPPPtfqx7rBUCvCybXWSOXi/xVcpZvPHrEYMdquszqme4XT6fqKVzqRDDdWKIHFtMqC/8Pd3qDTG1ncstxblc4HxeqVGIAYDp13bErwcvYlJyFCb0b2/QwqFUCFwqLKt5f6mwHK/8cAj/uyveth8ElS1HL645gL3n8i26z89DhUExITZ/PpEruPpKTC2MlHKtlHK6v7/pfm8ial5qlQJP3BBb8z7YR42XRnXFxqeH4LXbu+PGbmHwVitN3ptnZtBuu0AvbHlmKGYOiTbqoukdFYgZ1135R01ca19MGdDBpp8hu6gc//hmH26dv9niAAMAd/eJrHeRPCKq5FItMUTU/KSU+PtYFvpEBcLTRCAZ1SMcy7afwbUxwbh/QIeaWUl394nE3X0iUabRYUPaJaxOPY8NaZdrxrzklZjvrlEpzYeCOcNjsOl4Fvp3CsKjw2PgrjIdkuqj1emxbPsZvL3uCArKjAcSe7gp0LNtAKqXdZF1NmwUojJUzRzClmKihmKIIaImsys9B6//cgQ7TuXgHzfF4eHrOhldo1AIfDm9n9lF3DzclLi5exvc3L0N8ks02Hk6B+lnz2JYF9MDcgEgs6AMrbzUJls43JQKfP1Qf7hdJejUZ+fpHLyw5qDRyr/VRnRvjX/d0hUR3A6AyK4YYojIgEanR0mFDqUVOpRUaCu/1+jQK7IVlCamNf968CI2Hcuquq7y+pJyHUqqvq8sp/K19kyhD/88jvF9IuHvZTwLp6Gr0Pp7uWF41zCkKXLRLtDL5DU/7ruAZ7/bj/F9I/GPm+JMXmNtgLlUUIb//pyG7/acN3m+U4g3Xh59DQbGBFtVPhFdHUMMkQtZtOkUjlwsRGZhWYOCRm37XrrB5LTf3em5NbN/LFFQpsXHf5/A02aCha2KyrV4cc1BfLv7HADgo79OYHBsCPp1DLK5bCklPt9yGm/9egTFFTqj895qJR4dHotJyVEc30LUiBhiiJycRqfH+dxSnMkpwZmcEgDAff3am7x21Z5zOHDedJdHfUordCZDjKlxLQ0xODYEI7q3serehtDq9Nh8PKvmvZTAE1/txc+PDrJpDZbc4go8+fVe/JFmeifr2xIi8MzNcQjz87D6M4ioYRhiiJxAXklFTUg5k1OCM9lXvs/IK0WtPQMREeBpNsREBnpZHWJKTLQ4AICXBSFGpRDo0yEQjwyNQf9OtreIXE2Alxpv39UT9y7cDgBQKxWYnBwFHxsWrdt5OgezV+zBhfwyo3NxrX3xyphr0KeD7fsoEVHDMMQQOahnvt2HAxn5OJNdYnK2izkX8ktRodWb7MYwN26kNoWoXJ3WU62El1oJT7fKV3Or/Cd3CsZLo7rCS62Cl3v1PSp4Vd+vVsK7qjx3laJJd10eEB2MaYM64M8jlzH37gR0Dfezqhy9XuLDv07gf78dhU5vOK3I10OFJ2/ojHv7Rl51FhQR2R9DDFETyS/V4Gyt1pT07BJ4q5V4bqTpPXIOXyiwqtVELysXfusQ7G107oauYWjj54GIVl7wcbdP0Lgmwt/snkNNQUqJMo3pcTwA8OSNnfHEDZ3h4WZdt9flwnI8/lUqNh7LMjqXEBmA9+9JQNtW9YdDIrI/lwoxXLGXGpNOL5GRV2oYVHJKat6bWoyttZ+H2RATGeTd4MXSQn3d0T7IC+0CvRAZ6GW2iyexfSAS27ec7o6CMg2e++4ATlzIwZqucSZbQqxd9wUANh/PwpyVqcgqKjc69+Dgjnjyhs42Tc0mItu4VIiRUq4FsDYpKWlac9eFWpY7PtyCvWfzoK3T1VCfiwVlKNPoTLYSRAZeWVPEXaVAu0AvtA+8ElQiA73QPsgLbVt5WT241pntPpOL2Sv24FxuKQBg/oYTmDM8xi5la3V6vPfHMby/4bjRonSB3pVjbYZ0DrXLZxGR9VwqxBDVR6eXuFhQVjVwtriqRaVy5k+wtxqfTu5t8j4ppcUBptq53FJEh/oYHR+b2A6DY0PRPsgLIT7uUJgblOKi5q8/XhNgAGDuH0cxIDoISVG2tTRdzC/D7JV7sONUjtG5Ph0C8d7dCWjtz5lHRI6AIYZcTlG5tmZ2z9k63T7nckug0ZkOIyG+7mbLjAz0wu4zeVf9XLVKgXatPGtaUdoFeqF9kDfC/EyX2yHY2+S4Fqr02h09cPPcv5FVVLndQGyYLwJMLJxniQ1pl/D4V6nIrdP1JwTwyNAYzB4azcG7RA6EIYacWplGh8IyLQrLNFWvld8rFAI3dmtt8p6hb/2JS4XGYxzqc7mwHKUVOpNdN5FVs36CfdQ13T6Rtbt+grwQ5uvB1hQ7CvF1x9t3xWPyZzswOs4fr4/vb/XgXY1Ojzd/PYIFf580+Tlzx8UjOZqr7hI5GoYYajZlGh0Kyir/xRvqa7p5/vVf0nC5sNwopFR/b2512batPM2GmMhAL6tCDACczS1BbJiv0fEHBnXEg4M71WxWSE1jcGwIfntsMLQ556wOMGdzSvDIij1IPWvckjYoJhj/uyv+qq1wRNR8+BuX7KJ6Z+HCMi0KyjQoMAgbVw8gw7uEYuEk02NNVu0+h8wCywNHQanxTKBqkUFeSEnPNXs+0Ftda/CsJ9oHele+D/JCazOrsPp72taNQaalZxfjxe8P4vU7ephdATc61AdpxsNXGmT7yWxMW5JitA6PUiHw+PWxeHhwJ7aeETkwhhgnIaWETi+h0Ulo9HpodRJanR4Vuqrv9frKc7rKV61Oj6hgb5O/+LOKyvHLgYvQ6vTQ6uWVMnR6VFS9avUS5y9lQ7GzyCCAjOsdaXLn4dIKHR5ettuqn+1qC7n5erhZFWKKyrWQUppc76RTiA+igqrHpHgZjFGJDPSCrw1L0pP9/LjvAp75dh8Ky7V4dGUqvpja1+QGlNZan5aJh7/YjXKtYWteG38PvHdPAnrbOECYiBofQ4wd/XLgAnafyYOmKhTUBIqq0FEZFioDQu2wodFJzL07HjEmuilW7jiD59ccMDvY9Gr+77buGN830uh4Rl4pnlt9oIGlFBq8u5hfavIqHw/rH6XCq4aYq5frphTw9XCDr4eq8su9+ns3aHQSapXxX3ozh0Rj5hCuFeTI1qdlYubyK6F468lsfLDhOB4ZZp8p1GtSz+OJr/YazSgbFheKt+7siVbeart8DhE1LoYYO/r7WBaWbz9j1b1F5ab/IhcCVgUYANDqTY8XUSmsn11hrtXETamAp5sSpRrT++uYroeAr4cKflcJKtMHdUR+qcYwqHi4wc9DBT9PtyZfxp6axuDYUPTvGIStJ7NrjqVdLDTbumaJL7al4/k1B4zWf3ni+ljMGhrN54nIiTDE2JHahqmX5tYYsSVwmAs/bkrrf0kXlpkfazKyRxvoJWqCyZXgYRxAfD3c4OFWfwC5uRF3OSbHpVQIzL07HiPe24iCUi3+dUsXTOzf3uaA8cGfx/HGL0cMjgkB/HvMNWY3zSQix8UQY0cqG/rrNWZm2ahqBQ6FAFRKBdRKBVRKAZVCATelgFvVezeFAm6qK8fNzagI8FLjnj6RUCsFVFX3qpUKqBRV5VSVnZ9zGbFR7QwCSOBVmtnfvLOn1T8/UV2hfh6YN74XvNRK9GgbYFNZUkq89ksaPv7LcAq1SiHw9l09MSY+wqbyiah5uFSIaey9k4bGhSLY1x0qRa1goVTUhIIrgUMBN4WAm0pRc22UmUXNbuneBjdd0xpuCoXdZkmE+Lrjv7d3r/e6tLRyxMWxJYQaz8nLRcgurjA7iLZfxyCbP0Onl3hu9QGs2GHY1euuUuDD+3phaFyYzZ9BRM3DpUJMY++dlBwdbPcFsVRKhWv9TyKXsSb1PJ5dtR+eaiV+mj0IoWamUNuiQqvH41+l4od9FwyO+7ir8OmkJPS1Q0gioubD9bOJqElJKfGv7/ZjzspUFFfokFVUgTkrU6Gzcu8pc0ordJi+NMUowAR6q7FiWj8GGKIWgCGGiJqUEALBPobjtbaezMYvBy7a7TMKyjSYuGg7/jxy2eB4G38PfPVgf3Rv62+3zyKi5sMQQ0RNbvawGPSvaglRqxR49bZrMKK76W0iLJVVVI57FmzDztOGqzJHBXnh64f6m9wxnIicE4dbEFGTq55CPXP5brw0uhu6hdveMlJUrsWpy8WYs3IPTmYVG5zr0sYPS6b04R5IRC0MQwwRNZoyjc7sxoyhfh74+qFkm8r//VAmPvzrBA5n5KNEc8LkNYntW2HR5N7c34qoBWJ3EhE1ik3HsjDojQ3YXmvVXXtaui0dU5ekYFd6Lko0ptdZGhQTjKUP9GGAIWqhGGKIyK70eon5G45j4qLtuFxYjpnL9yCzoMzq8vJKKnAxvwyy1j4BS7eexvP17P81ontrLJyUBC81G5yJWir+6SYiu9p7Lg9v/nplaf+sonLMXLYbK6b3g5uFW3Ms3HgSr/+SBo1OItjHHYntAxDi644vthnvUaZWKtDa3wPhAR4Y3TMC43q3s+uu10TkeBhiiMiuEiJb4ZGh0Xh//XEAlXsTDY4NgdLCfY8WbzmN//x4uOZ9VlE5fj2YaXSdWqnAs9eFYuKwXnZb1ZqInANDDBHZ3aPDY5F6Ng8Hzudj7t0JuDY2xKL7V+0+hxe/P1jvdWqVAgsmJKK1zGGAIXJBDDFEZHeVU6gTUFKhRdtWXhbdu+7gRTz1zb56r1OrFPhkYhIGx4YgLS3H2qoS9jf5fQAAIABJREFUkRNjiCEiq2TklWLHqRzcmmB6B+hAb/VVdz03ZfPxLMxavsdgCwI3pcBH9yUi1NcDu9JzsOdsHrR6iemDOqJnO9t2tyYi58YQQ0QW23QsC7NX7kFeSQVCfd3tsvHpnjO5mLYkBRW6K9OlFQJ4d1wChnWp3Gm6e1t/TLb5k4iopXCpKdZCiFFCiAX5+fnNXRUip/XZ5lOYuGg7cooroJfAIyv24EJ+qU1lpl0swOTPdqKkQmdw/LXbe+CWHm1sKpuIWi6XCjFSyrVSyun+/tz8jchaUcHeqL3fdHZxBZZuTbe6vNNZxZjw6Q7kl2oMjj93Sxfc1bud1eUSUcvnUiGGiGw3pHMo5gyLAVA5ffqJ62Px5A2drSrrQn4p7l1YuShebbOHRmPqoI4215WIWjaOiSEii80eGoMz2SW4NSHC4unT1bKLynHfwu04n2fYFTU5OQqPXR9rj2oSUQvHEENEJlVo9VAphMn1VxQKgf+Ni7e67IIyDSZ9tgMnLhvuNn17rwi8MLIrhIUL4xGRa2J3EhEZOZ9Xijs/3ooP/zK9M7QtSit0mPp5Cg6cLzA4fkPXMLxxRw8uWkdEDcaWGCIysOlYFh5ZsRu5JRrsP5eHHm39MSjGui6juiq0ejy8bBd2nDZcnG5gdDDeH58AlYV7KxGRa+NvDCKqUVKhxaNfpiK3pHKmkF4Cs1fsMRq3Yg2dXuLxr1Lx55HLBscTIgPw8YREuKuUNn8GEbkWhhgiquGlVuGdcT1Re0hKr8hW8FHb1mir0enx7Kr9+GHfBYPjca198fnkPvB2Z6MwEVmOIYaIDAyKCcET18dCCODJG2LxycQk+Hu5WV3e+bxSjPt4K75MOWtwPCrIC0se6GNT2UTk2vjPHyIyMuO6aFwbG4IebW3bm+j3Q5l48pu9yCsxXMiujb8HvpjaF6G+HjaVT0SujS0xRC6oXKvDkq2noddLk+cVCmFTgNHo9Hj1x0OYuiTFKMC09vPA0gf6Wry7NRFRXWyJIXIx5/NKMWPZbuw9m4ec4go8Oty+C8udyy3BIyv2YM+ZPKNz13UOwf/uird4d2siIlMYYohcyK70HExbsgs5xRUAgLl/HEN8uwBc1znULuWvO3gRT32zz2gfJKVC4KkbO2P6oI5cB4aI7IYhhsiFhPh4QC+vdCFJCfzvt6MYHBti0yq5FVo9Xv8lDZ9uOmV0ro2/B96/JwFJUYFWl09EZArHxBC5kMggL3xwby8oq1pDhsWFYumUvjYFmLM5Jbjz460mA8zQuFD8NHsQAwwRNQq2xBC5mOROwXhpdDfkFldg1pBom7p3fj14EU99vRcFZVqD4yqFwNM3dcbUgew+IqLGwxBD5IIm9Gtv0/0VWj3++/NhfLb5tNG5iABPvHdPAhLbt7LpM4iI6uNS3UlCiFFCiAX5+fnNXRWiRrXu4EU8/mWq2SnUtjiTXYKxH20xGWCGdwnFj7MHMsAQUZNwqZYYKeVaAGuTkpKmNXddiBqDlBLzNxzHW+uOAgDatvLE4zd0tlv5vxy4gKe+2YdCE91Hz9wchwcGdrBpfA0RkSVcKsQQtXTv/HYU760/XvP+vfXHEdvaFyN7hNtUbrlWh//78TAWb003OhcR4Il54xOQEMnWFyJqWi7VnUTU0o3rE4lgnysLySkVAgWl2qvcUb/07GKM/XCryQBzfdcw/DR7EAMMETULhhiiFiQiwBMf3ZcIN6VAgJcblj7QB+P7Rlpd3o/7LmDke5uw/7zhODI3pcALI7tiwYREbuBIRM2G3UlELUxSVCDeuzsBXcP90D7I26oyyjQ6vPrjYSzdZtz60raVJ+aP74We7WzbHJKIyFYMMUROSKvTo1yrh7e76T/CN3dvY3XZp7KKMWv5bhzMKDA6d1O31nh9bA/4e7L1hYiaH0MMkZPJL9XgkRV7oBDAp5N616y+aw9r92bgn6v2o6jccByNWqnAsyPiMCk5irOPiMhhMMQQOZETl4swbXEKTmYVAwBe/yUNz47oYnO5ZRod/v3DISzbfsboXGSgF+aP74Xubf1t/hwiIntiiCFyIi+vPVQTYABgwd8n0S3cD2PiI6wu8+TlIsxcvgeHLxh3H43o3hqv3dEDfh7sPiIix8PZSURO5M2xPRDm517zPiEyAP07Blld3prU8xj1/iajAKNWKvDvMd0wf3wvBhgiclgMMUROJMzPAwsmJEGtUuD2XhFYMa0fQv08LC6nTKPDP1ftx5yVqSiu0Bmcax/khVUzkjGhP8e/EJFjY3cSkZPp2S4AP80eiE4hPlaFjBOXizBz2W6kXSw0OndLjzZ47fbu8GXrCxE5AYYYIgd04Hw+IgI80cpbbfJ8dKivVeX+tP8Cnvx6L0rqtL6oVQq8MLIr7u0bydYXInIaDDFEDmbt3gw89c1eJLRrhSUP9IGb0vZeX51e4n+/HcH8DSeMznUI9sa88QnoFs7ZR0TkXDgmhshB6PUSb/16BI+s2IMyjR5bT2bjPz8csrnc/FINpi1JMRlgRvcMx9pHBjLAEJFTYksMkYPQ6iW2n8o2OLZ4azpGx0cgsb11Gywev1SI6Ut2GUzLBqr2PhrVDfex+4iInBhbYogchFqlwIf3JSIiwLPyvVKBN8b2sDrA/HYoE7fO32IUYIJ93LFiWj9M6NeeAYaInBpDDJEDCfZxx4KJiYgM9MKK6X1xV1I7i8vQ6yXm/n4M05akGG0f0LNdAH54ZCCSogLtVWUiombD7iQiB9Mt3B9/PDHYqgG9hWUaPPHVXqw7lGl0bmxiW/zn1mvg4aa0RzWJiJodW2KImphGp8dbvx7B5cJys9dYE2BOXi7CbR9sMQowSoXAy6O74c2xPRhgiKhFYUsMURPKLa7AjGW7sfVkNradzMbyaf2gVtn+b4kNaZcwe+UeFJYZdh8Feqsxf3wv9O9k/dYERESOii0xRE0kI68Uo+dvwtaTlTOQUtJz8fzqA5BSWl2mlBIf/HkcUxbvNAow3cL98P2sAQwwRNRisSWGqImE+rojMtALZ3NKa45tOp6F3BINAs2szHs1JRVaPPX1Pvy4/4LRuVvjw/Hf23vAU83uIyJqudgSQ9REVEoF5o/vhfZBXgCAPh0C8f2sAVYFmDPZJbj9gy1GAUYhgOdu6YJ3xsUzwBBRi8eWGKImFOClxsKJSVi+4wz+eXMXq8bDbDqWhVkrdiOvRFOnbDfMu6cXBsYE26u6REQOjSGGqBFIKc0uJBcT5osXR3WzqsyFG0/hvz8fhr7OMJq41r5YMCEJkVWtPERErsClupOEEKOEEAvy8/ObuyrUgu1Kz8WoeZuQWVBmtzJLK3R49MtUvPqTcYC5pXsbrJqRzABDRC7HpUKMlHKtlHK6vz83u6PG8c2uc7hnwTYcOF+A6UtSUKbR2VzmudwSjP1oC9akZhgcFwJ4+qbOmDc+AV5qNqoSketxqRBD1JjWp2Xiya/3okKnBwDsPZePf67ab9MU6q0nsjF63mYczCgwOO7rocKiyb0x47po7n9ERC6LIYbITgbHhmJI5xCDY238PWBNhpFSYvGW07jv0+3IKa4wOBcd6oPvZw3EkM6htlSXiMjpMcQQ2YlSITD3ngR0CvGGu0qBuXfH4+mb4qBQWNZSUqbR4elv9uHF7w9CV2cAzA1dw7B65gB0CPa2Z9WJiJwSO9KJ7MjPww0LJ/VGQakGPdsFWHz/hfxSPPTFbuw9m2d07rHhsXhkaLTFoYiIqKViiCGykJQSmQXlaO3vYfK8ta0kKadz8NAXu5FVZLgxpI+7Cu+Mi8f1XcOsKpeIqKVidxKRBcq1lV09I9/fiPN5pfXf0EDLtqfjnk+2GQWYDsHeWD0zmQGGiMgEhhiiBrpcWI7xn2zH17vOIauoAlMXp6CkQlv/jVdRodXjn6v241/fHYBGZzj+ZWhcKFbPHIDoUF+bPoOIqKViiCFqoFW7z2FXem7N+8MXCvDCmoNWl3epoAz3fLINK3acMTo3a0g0Fk5Mgr+nm9XlExG1dBwTQ9RA0wZ1xK70XKw7lAkAiArywkODO1pVVurZPDy4NAWZBYbdR15qJd66sydGdG9jc32JiFo6hhiiBlIoBN4ZF487PtyCEF93zLunF/y9LG8p+SrlLJ777kDNonjVIgO9sGBiIuJa+9mrykRELRpDDJEFvN1V+GJqXwR4ukGltKw3VqPT4z8/HMLirelG5wbFBOP9exIQ4KW2V1WJiFo8hhiiOs7mlCCzoAxJUYEmzwf7uFtcZlZROWYs240dp3KMzj14bUc8dWNni0MREZGrY4ghqmXbyWzMWLYbeimxZuYAtA+yfWXc/efy8eDSFGTkG+5q7eGmwOt39MCY+AibP4OIyBXxn35EVZZvP4P7FlbuVZRXosHUxSkoLNPYVObqPecx9qMtRgEmIsAT3z6czABDRGQDhhiiKrklFdDW2qvo2KUik9OfG0JbNf7l0S9TUa41HMDbv2MQ1j4yEN3C/W2qLxGRq2N3ElGVGdd1QtrFQqzdmwGgcq+iqQMtn0KdW1yBWSt2Y/PxbKNz9w+IwrMjusCN41+IiGzGEENURQiBN+7ogcyCMtyfHIWbrVir5VBGAaYvTcG5XMMtCdQqBf7vtu4Ym9jWXtUlInJ5DDFEtXiqlfhyej8IYflO0T/sy8BTX+9DqUZncLy1nwc+npBo1a7WRERkHkMMuRQpJT786wS6tPHDkM6hJq+xNMDo9BJvrTuCD/88YXSud1QrfHBvIkJ8LZ+WTUREV8cQQy6jTFO5A/X3ezPg667CdzMHIDrUx6Yy80s0mL1yD/46etno3H39IvHCyG5Qqzj+hYioMfC3K7kErU6Puxdsw/dVg3YLy7WYtiQF+SXWT6E+mlmIMfM3GQUYN6XAa7d3x39u7c4AQ0TUiPgbllyCSqnAyB6GA3VziitwOrvYqvJ+OXARt83fjNPZJQbHQ33dsXJ6f9zdJ9LquhIRUcOwO4lcxgMDOyDtYiG+2XUOMaE+WDgpyeIVefV6iXf/OIb3/jhmdC6+XQA+npCIMD8Pe1WZiIiugiGGXIYQAq/edg1Cfd3x8HWd4Oth2Q7UhWUaPPZlKn4/fMno3F1JbfHvW6+Bu0ppr+oSEVE9GGKoxZFSmp1h5K5S4umb4iwu88TlIkxfkoITlw27n1QKgRdHdcV9/dpbNS2biIisxzEx1KJkF5Vj3MfbsCHNuLXEWuvTMnHrvM1GASbYR41lU/tiQv8oBhgiombAlhhqMY5lFmLK4p04m1OKg8vz8c3DyejSxs/q8qSUmL/hON7+7SikNDzXPcIfH09IRHiAp421JiIia7ElhlqE/FIN7vx4K87mVC73X1yhwwOf78SlgrJ67jStuFyLGct24611xgHm9oQIfP1QfwYYIqJmxhBDLYK/pxseHRZjcCwq2Nuqgbbp2cW47YPN+PnARYPjSoXACyO74u27esLDjQN4iYiaG7uTqMWYPKADTmUVY/HWdNzTpx1eGXONxbtF/3X0Mh5ZvhsFZVqD46283DBvfC8MiA62Z5WJiMgGDDHUojw/siv6dAjCiO6tLRpsK6XEgr9P4vVf0qCv033UpY0fFkxIRLtALzvXloiIbMEQQ06ntEIHT7Xp7hyVUoFb6qzM25Dynv52H9ZWbUlQ28gebfDG2B7wUvOPChGRo+GYGHIqB87nY+jbf+KXAxfsUt7ZnBLc8eEWowCjEMAzN8fh/XsSGGCIiBwUfzuT0/jtUCZmr9iDUo0Oj36Zii/9PdGzXYDV5W05noWZy3cjt84mkH4eKrw/vhcGx4bYWmUiImpEbIkhp5BZUIaZy3ejVKMDAJRp9Ji6JMWqKdRSSny66RQmLNphFGBiw3zw/ayBDDBERE7AaUOMEOJaIcT3QojzQggphJjc3HWixhPm54F/j+lmcGx0z3AE+bhbVE6ZRocnvt6Lf/9wCLo6I3hv6tYaq2YMQFSwZZtCEhFR83Dm7iQfAAcALKn6ohZuXO9InMwqxsKNp/Dy6G64r197i+7PyCvFQ1/swr5z+QbHhQAeHx6LmUOioVBw+wAiImfhtCFGSvkTgJ8AQAjxefPWhprKP26Mw8ju4eje1t+i+3acysGMZbuQVVRhcNzXXYV3xsVjeNcwe1aTiIiaQIO7k4QQY4UQ7wshNgohCqq6cL6o5562QohFQogMIUS5EOK0EOJdIUQr26tOLdWF/FKz5xQKYVGAkVJi6bZ0jP9km1GA6RjijdWzBjDAEBE5KUvGxDwHYBaAeADn67tYCNEJwC4A9wPYAeAdACcBzAGwVQgRZHFtqcVbk3oeg9/8E6v31PuI1Uur0+P5NQfw/OoD0NYZ/zK8SyhWzxyATiE+Nn8OERE1D0tCzGMAYgH4AXi4Add/ACAUwGwp5a1SymeklENRGWY6A3i19sVCiP9Ute5c7es6C+pLTkRKiXd/P4o5K1NRodXj6W/2YefpHKvLyy/V4P7Pd+KLbWeMzs0eFoMFE5Lg5+FmS5WJiKiZNTjESCk3SCmPSVl3T19jVa0wNwA4DWB+ndMvAigGMEEIUXsayLsAutTztaOh9SXncuJyMT7YcKLmfYVOjweX7kJ+qeYqd5l2JrtyAbuNx7IMjnurlfjovkQ8fn0sB/ASEbUAjTWwd0jV6zoppb72CSlloRBiMypDTj8Af1QdzwJg+LcOuYzoUB+8eWcPzFmZCgBQKxV4fmQX+Hta1lqScjoH05fuQk6x4fiXiABPfDo5CXGt/exWZyIial6NFWI6V70eNXP+GCpDTCyqQoylhBA+AKKr3ioARAoh4gHkSCmN+hCEENMBTAeA8PBwpKWlWfOxLiUrK6tJ/zt19gDui2+F7w/n44WhrdHFs8iiz//jRCHe3XwJGr3h8bhgd7w4LBTIy0BanvH+SOT8mvpZJbIWn1X7aqwQUz19JN/M+erj1q8ZDyQB2FDr/ctVX4sBTK57sZRyAYAFAJCUlCTj4uJs+GjXkJaWhqb+7/TvzhKPFJYjzM+jwffo9RLv/H4U72+8ZHRuVM9wvDm2BzzcTG8YSS1DczyrRNbgs2pfzrxOzJ8AOLDBCe05k4uESNOz7IUQFgWY6hV4f9xnvCHk7GExeGx4DITgY0JE1BI11rYD1S0t5hb0qD6e10ifTw5Ip5f4v58O47YPtmDFDuNZQ5a6VFiGcQu2GQUYtVKBd8fF4/HrYxlgiIhasMZqiTlS9Rpr5nxM1au5MTPUwpRUaDFnZSp+O5QJAHh+9QFEBnphQHSwVeUdvlCABz7fiYx8ww0gg7zVWDAxEYntA22uMxERObbGaompHqtygxDC4DOEEL4ABgAoAbCtkT6fHMzprBJsPHa55r1WL/H4V6koq9qV2hLr0zIx9sMtRgEmJtQHq2cOYIAhInIRjRJipJQnAKwDEAVgZp3TLwPwBrBUSlncGJ9vjhBilBBiQX6+ufHG1Fi6hvvh3XEJqO7dCfV1x8KJvS0acCulxKJNpzB1cQqKKwzDz7WxIfh2RjLaBXrZs9pEROTAGtydJIS4FcCtVW9bV732r7X5YpaU8slat8wAsAXAe0KIYQAOA+iLyjVkjgL4lw31toqUci2AtUlJSdOa+rMJuOma1vjHTXFYk5qBRZOT0Mbfs8H3anR6vPT9QSzbbjyWZmL/9nhhZFeolI3VsEhERI7IkjEx8QAm1TnWseoLANIB1IQYKeUJIUQSgFcA3ARgBIALAOYCeFlKmWttpcl5PXhtR0xOjrKoBSa/VINZy3cbrcCrEMALI7ti8oAO9q4mERE5gQaHGCnlSwBesqRwKeVZVG4ASS5Co9PjlwMXMbJHG5Mzg4QQFgWYM9klmLJ4J45fKjI47uOuwvvjEzCkc6jNdSYiIufktOvEkOOp3WKSVVSO+21sIdl5OgfTl6Qgt8Rw/6SIAE8smtwbnVv72lQ+ERE5N4YYsouzOSW4//MrLSb//uEQ2gd5YWhcmFXlrdp9Ds98ux8VOsM9BBIiA7BgQhJCfN1trjMRETk3joQkuygq1+JCXmnNe70EXv/5CHT6ejc9N6DXS7z16xE8/tVeowAzqmc4VkzrxwBDREQAGGLITrq08cO88b2gqBoGkxAZgC+m9oVS0fAVc8s0OjyyYg/mbThudG7OsBi8d3c890AiIqIaLtWdJIQYBWBUdHR0vdeS5YbEheLFUd2w83QO3rqzp0WB41JhGaYt2YW9Zw13olCrFHhzbA+MiY+wd3WJiMjJuVRLjJRyrZRyur+/uS2dyFaTkqPw/j0JFgWYQxkFuHXeZqMAE+StxoppfRlgiIjIJJcKMWS77KJyfPzXCUhpfqyLJZsu/nE4E3d+ZLyFQGwYtxAgIqKrc6nuJLLN8UtFmPL5TpzJKYFOSsy4zvpuOSklFm0+jVd/PIS6Y38Hx4bg/fEJ8PNws7HGRETUkjHEUINsP5mNqUtSUFimBQC88csRRAV5Y0T3NhaXpdHp8eL3B7HcxBYCk/q3x/PcQoCIiBqAf1NQgwT5uKNuJ9G3u85dtVvJlPxSDe7/bKdRgFEI4OXR3fDymGsYYIiIqEH4twU1SHSoDz66LxGqqinT45La4aMJiRaNf0nPLsbtH2zGpuOGeyD5uKuwaHJvTEqOsmeViYiohWN3EjVYcnQwXr3tGuSWaPDgtR0tCjA7TuXgwaXGWwi0beWJTydxCwEiIrIcQwxZZFzvSIvv+XbXOTyzah80OsOup16RAVgwMQnBPlyBl4iILOdS3UlCiFFCiAX5+fnNXRWHdSijAP/54ZDFY11Mqd5C4Imv9xoFmNE9w7F8Wj8GGCIisppLtcRIKdcCWJuUlDStueviiDYcuYRZy3ajuEIHL3cVRrSzvqzSCh2e+DoVP+2/aHTuseGxmD0s2qLuKCIiorpcKsSQeWtSz+Pxr/bWbNj43h/H4DYwBHFxlpd1qaAM05akYO85wxYvtUqBt+7sidE9w+1RZSIicnEMMQQAiG8XAH9PN+QUV9QcO5VbcZU7TDuUUYCpi3carcAb7KPGxxOSkNi+lc11JSIiAlxsTAyZ1z7IG59MTIRapYAQwIujumJ672CLyvj9UCbGmthCoHOYL76bMYABhoiI7IotMVQjsX0g3rkrHmqVAtd3DUNaWlqD7pNS4tNNp/DqT4dRdzzw4NgQzBufAF9uIUBERHbGEEMGbulh2TYCGp0eL6w5iBU7jLcQmJwchedu6cIVeImIqFHwbxcXs+V4Fp76ei/0dXddtEJ+iQaTP9thFGCUCoFXxnTDS6O7McAQEVGjYUuMC/k65Sz+uWo/tHqJVt5qPDuii9VlpWcXY8rnO3HicrHBcV93Febd2wuDY0NsrS4REdFVMcS4iKXb0vH86gM17xf8fRKRgV64r197i8vafjIbD32xy+QWAosm90ZsGLcQICKixudSbf2uvGLv0LhQhPgaro5bUqG1uJxvd53DfZ9uNwowvSIDsHrmAAYYIiJqMi4VYqSUa6WU0/39/Zu7Kk0uIsATiyb1hqebEmqVAvPH98L0azs1+H69XuLNX9NMbiEwJp5bCBARUdNjd5IL6d7WH/PvTYC/p9qiNVtKK3R4/KtU/HyAWwgQEZHjYIhpgaSUZkPF0Lgwi8q6VFCGqUtSsI9bCBARkYNxqe4kV7Am9Twe+mIXtDq9zWWdyC7HmPmbjQJMsI8aK6f3Y4AhIqJmxRDTQkgpMW/9McxZmYpfD2bi5bWHIOsun2uB3w5l4omfz+OCiS0EVs8cgF6R3EKAiIiaF0NMC/HBnyfw1rqjNe+XbkvHp5tOWVyOlBILN57E9KUpKNMahqDrOofgm4f7o20rL5vrS0REZCuGmBbizsS2iAjwrHmvVioQ5KO2qAytTo8Xvz+I//xovAfS5OQoLJyYxD2QiIjIYTDEtBChfh5YNLk3fN1V8Pd0w9IH+uC2hLYNvr+4XIsHl+7Ckq3pBseVCoF/cwsBIiJyQJyd1IJ0bu2LjycmIszPA51CfBp8X2ZBGaZ8vhMHMwoMjnu5CXw4oTe3ECAiIofEEOOEdHoJpcL0FOrkTsEWlZV2sQBTPtuJjDoDeMP9PfD8dcEMMERE5LBcqn/A2bcdkFJiwd8nMHHRdlRobZ9CvfHYZdz54VajANMt3A/fzRyADq24Ai8RETkulwoxzrztgFanx79WH8D//ZSGzcez8ex3+22aQv3lzjO4/7OdKCw33D9paFwovnqwP8L8PGytMhERUaNyqRDjzN7+7SiWbz9T8/6bXecwf8Nxi8uRUuKtX4/gH9/uh1ZvGIIm9GuPBRMS4e3OXkYiInJ8DDFOYtqgjmgfdGV9Fl93FeLbWbbgXLlWhzkrUzGvTvgRAnjuli54ZQxnIBERkfPg31hOItBbjc8m94a/pxsiAjzx7YxkDIxp+CDe3OIKTFi4A9/vzTA47q5S4IPxvTB1UEdu4khERE6F/QZOpGOIDxZP6YPwAA+E+jZ8zEp6djEmf7YTp7KKDY4HeavxyaQkbiFAREROiSHGAZVpdPBwU5o8F98uwKKydqXnYtqSFOQUVxgc7xjijc8n90FkELcQICIi58TuJAei00u8vPYg7l24HWUanc3l/bjvAu75ZJtRgOnTIRCrHk5mgCEiIqfGEOMgSioql/3/bPNp7ErPxZNf74Veb90UaiklPv7rBGYu3220nsyt8eFY+kAfBHhZtq8SERGRo2F3koN445cj+P1wZs37H/ZdQMcQHzx+faxF5VRv4ris1nTsarOHRuOx62M5gJeIiFoEtsQ4iMeuj0V06JX9jsL83HFjtzCLyigq12LqkhSjAKNSCLwxtgcev6EzAwwREbUYDDEOwt/TDZ9N7o1gHzW6tPHD6pkD0C284SsLX8wvw50fbcWfRy4bHPd1V2HxlD64K6mdvatMRETUrNid5EDaBXph2dR+iGjlCR8LVs09lFGAKZ/vxMUCwz2QIgI88dn9vRGEJ63oAAAMjElEQVQb5mvvqhIRETU7hpgmptdLFJZr4e/pZvJ859aWBY4/j1zCzGW7UVxhOJupe4Q/Pp2UhFDugURERC0Uu5OaUJlGh0dW7MF9C7ejpEJb/w31WL79DB5YnGIUYIZ3CcWXD/ZjgCEiohbNpUKMEGKUEGJBfn5+k392dlE5xn+yDT/uv4D95/Mxe0UqdFZOodbrJV77OQ3PfrffqIzJyVH4eEISvNRsZCMiopbNpUKMlHKtlHK6v3/DB8zay9u/HcXuM3k1738/nIl3fz9qcTllGh0eWbkHH/11wuC4EMALI7vipdHdoFRwBhIREbV8/Od6E3l2RBeknsnDoQsFAICYUB+LZwzlFFdg2pIU7ErPNTju4abA3LsTcGO31narLxERkaNzqZaY5uTjrsKiyb3R2s8DA6OD8c3DyWgX2PBl/09lFeP2DzYbBZhgHzVWTu/PAENERC6HLTFNqLW/B75+qD9a+3vATdnw/LjzdA6mL0lBbonG4Hh0qA8+m9zbojBERETUUjDE2Fm5Vof8Ug1CfU3PDLI0cKzdm4EnvtqLCp3hHkj9Ogbi4/uS4O9leqo2ERFRS8cQY0d5JRWYvnQX8ks0+Prh/vDzsD5gSCnx4V8n8MYvR4zO3Z4Qgdfu6AG1ir2BRETkuvi3oJ2kZxfj9g+2YMepHBzJLMTMZbuhqdN60lAanR7PfrffZIB5dHgM3r6rJwMMERG5PP5NaCcf/nkCJ7OKa95vPJaF9/84ZnE5hWUaPLA4BSt2nDU47qYUePvOnnh0OHehJiIiAtidZDcvjuqGwxcLsfds5VowfaICcf+ADhaVkZFXiimf70TaxUKD474eKnw8IRHJnYLtVl8iIiJnx5YYO/FUK7FwYhLatvLEmPhwLJ3aB6281Q2+/2BGPm77YLNRgGnbyhPfzUhmgCEiIqqDLTF2FOLrjlUzkhHi425Rl8+GtEuYtdx4E8eebf2xcFJvhPi627uqRERETo8hxs7MTa02Z+m2dLy45gDqbqN0Q9cwzL07AZ5qpR1rR0RE1HIwxDQTvV7itV/SsODvk0bnpgzogH/d0oV7IBEREV0FQ0wzKNPo8PhXqfhp/0WD44qqTRwnWzggmIiIyBUxxDSx7KJyTF2Sgj21drQGAE83Jd67JwHXdw1rppoRERE5F4aYJnTichHu/2wnzuSUGBwP8XXHokm90b2tfzPVjIiIyPkwxDSR7SezK7ckKDXcxDE2zAeLJvdG21bcxJGIiMgSDDFNYE3qeTz19T6jTRwHRAfhg3sT4e/JTRyJiIgs5VIhRggxCsCo6OjoJvk8KSXmbziOt9YdNTp3Z2JbvHpbd+6BREREZCWX+htUSrlWSjnd37/xx55odHr849t9JgPME9fH4o2x3IWaiIjIFi7VEtNUCso0mPHFbmw6nmVwXK1U4I2xPXBrQkQz1YyIiKjlYIixs/N5pbj/sx04mllkcNzf0w0fT0hEv45BzVQzIiKiloUhxo72n8vHlMU7cbmw3OB4u0BPfH5/H3QK8WmmmhEREbU8DDF28vuhTDyyYg9KNYabOMa3C8DCSUkI9uEmjkRERPbEEGMHS7aexkvfHzTaxPHma1rjnXHx8HDjJo5ERET2xhBjo+OXivDy2kNGAWb6tR3xzE1xUHATRyIiokbBOb42ig71wStjutW8Vwjg32O64dkRXRhgiIiIGhFbYuzg3r7tcSanBEu3pmPe+AQMjeMmjkRERI2NIcZO/nFjHO7pHYmoYO/mrgoREZFLYHeSnSgUggGGiIioCTHEEBERkVNiiCEiIiKnxBBDRERETokhhoiIiJwSQwwRERE5JYYYIiIickoMMUREROSUGGKIiIjIKTHEEBERkVNiiCEiIiKnxBBDRERETklIKZu7Dk1OCHEZQLqJU/4A8m0o2tr7Lbmvodc25Lr6rgkGkNXAejk6W//fOtrn8lk1xGfVcT+Xz6ohPqv1ay+lDGnQlVJKflV9AVjQHPdbcl9Dr23IdfVdAyCluf+fOMr/W0f7XD6rRuf5rDro5/JZNTrPZ9WOX+xOMrS2me635L6GXtuQ62z9eZ1Jc/2sjfW5fFZbLj6r9rmfz2rja/af1SW7k6hhhBApUsqk5q4HUX34rJKz4LNqX2yJoatZ0NwVIGogPqvkLPis2hFbYoiIiMgpsSWGiIiInBJDDBERETklhhhqFEKIa4UQ3wshzgshpBBicnPXiUgIMUMIcUoIUSaE2CWEGNTcdSKqi78/G44hhhqLD4ADAOYAKG3muhBBCDEOwFwA/wcgAcAWAD8LISKbtWJExvj7s4E4sJcanRCiCMAsKeXnzV0Xcl1CiO0A9kkpp9U6dgzAN1LKfzZfzYjM4+/Pq2NLTAshhBgrhHhfCLFRCFFQ1QT5RT33tBVCLBJCZAghyoUQp4UQ7wohWjVVvYmqNeYzLIRQA0gEsK5OEesAJNv3J6GWjr9vHYequStAdvMcgJ4AigCcAxB3tYuFEJ1Q2ZweCmANgDQAfVDZfHmTEGKAlDK7UWtMZKgxn+FgAEoAmXWKyQQw3F4/ALkM/r51EGyJaTkeAxALwA/Aww24/gNU/oGaLaW8VUr5jJRyKIB3AHQG8Grti4UQ/6n618bVvq6z749ELqZRn2EiO+Kz6iAYYloIKeUGKeUx2YBBTlX/KrgBwGkA8+ucfhFAMYAJQgjvWsffBdClnq8dNv4Y5MIa+RnOAqADEFbn2jAAF22pN7meJvh9Sw3EEOOahlS9rpNS6mufkFIWAtgMwAtAv1rHs6SUafV8lTTdj0AuzqJnWEpZAWAXgOvrlHM9Kpv5iRqLxb9vqeEYYlxT56rXo2bOH6t6jbX2A4QQPkKIeCFEPCqfs8iq95zOSvZgzTP8PwCThRBThRBdhBBzAYQD+KiR6kgEWPGs8vdnwzHEuCb/qtd8M+erjwfY8BlJAPZUfXkCeLnq+1dsKJOomsXPsJTySwCPonJQZiqAgQBGSCnTG6uSRLDu9y1/fzYQZydRo5BS/glANHc9iGqTUn6AykGWRA6Lvz8bji0xrqk6+fubOV99PK8J6kJkDT7D5Cz4rDYihhjXdKTq1dyYl5iqV3N9uETNjc8wOQs+q42IIcY1bah6vUEIYfAMCCF8AQwAUAJgW1NXjKiB+AyTs+Cz2ogYYlyQlPIEKpdbjwIws87plwF4A1gqpSxu4qoRNQifYXIWfFYbFzeAbCGEELcCuLXqbWsANwI4CWBj1bEsKeWTta6vuwz2YQB9UbmmwVEAyVwGm5oSn2FyFnxWHQdDTAshhHgJlas/mpMupYyqc087VE7ZuwlAEIALAL4D8LKUMrdxakpkGp9hchZ8Vh0HQwwRERE5JY6JISIiIqfEEENEREROiSGGiIiInBJDDBERETklhhgiIiJySgwxRERE5JQYYoiIiMgpMcQQERGRU2KIISKnIIQ4LYQ43dz1ICLHwRBDRE5PCHGnEEIKIe5o7roQUdNhiCGiluA2AKUAfm7uihBR02GIISKnJoRQA7gFwDopZUlz14eImg5DDBE5DFFplhDioBCiTAhxXggxTwjhf5XbhgLwQ+WOwNXlfF7VvRQlhHhQCLG/qrxMIcSCesojIiehau4KEBHV8i6A2QAuAFgAQANgDIC+ANQAKkzcczsALYC1Js69AeDGqnPrAAwBMA1ANCrDDxE5MYYYInIIQohkVAaYEwD6SClzqo7/C8AGAG0ApNe5R4HKkPNX9fV19APQXUp5pup6FYD1AIYIIfpIKXc01s9DRI2P3UlE5Cjur3p9tXYgkVKWAfinmXuSAYSiVldSHa9UB5iqsrQAPqt628e26hJRc2OIISJH0avq9S8T5zYB0Jk4fhsACWC1mTJTTBw7W/XayqLaEZHDYYghIkdRPdg2s+6JqhaULBP33AZgp5TyvJky80wc01a9Ki2uIRE5FIYYInIU+VWvYXVPVI1lCa5zLB5ABwCrGr9qROSIGGKIyFHsrnodbOLcQBi3nNxW9WpuPAwRtXAMMUTkKD6vev2XECKw+qAQwgPAf01cfxuAQ1LKo01QNyJyQAwxROQQpJSbAbwPoBOAA0KI94QQbwM4gMrlIC5UXyuEiAbQHWyFIXJpDDFE5EjmAHgEleNjHgRwD4BfAQyH4UJ31V1JHA9D5MKElLK560BEZBEhxBYA4VLKqOauCxE1H7bEEJFTEUK0QeVKvObWhiEiF8GWGCIiInJKbIkhIiIip8QQ8//t1gEJAAAAgKD/r9sR6AoBgCWJAQCWJAYAWJIYAGBJYgCAJYkBAJYC97XflPAsUkYAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 576x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "ds = results[\"ds\"]\n",
    "gammas = results[\"gammas\"]\n",
    "mdl_theory = results[\"mdl_theory\"]\n",
    "mdl_practice = results[\"mdl_practice\"]\n",
    "thm3_bnds = results[\"mdl_thm_3\"]\n",
    "mse = results[\"mse\"]\n",
    "eigsum = results[\"eigsum\"]\n",
    "tnorm = results[\"tnorm\"]\n",
    "\n",
    "# plt.figure(figsize=[5, 5])\n",
    "\n",
    "mpl.rcParams['legend.fontsize'] = 20\n",
    "lss = ['-', '--', '-.', ':']\n",
    "plt.figure(figsize=[8, 6])\n",
    "for j, d in enumerate(ds):\n",
    "    ns = np.asarray(d/gammas, dtype=int)\n",
    "    plt.plot(d/ns, mdl_theory.mean(0)[j, :], color=cs[j], linestyle=lss[j],linewidth=4.,  label='d=%d'%(d))\n",
    "plt.grid(True, alpha=0.5, linewidth=1.)\n",
    "plt.legend(frameon=False)\n",
    "plt.title('MDL-COMP', fontweight='bold', fontsize=20)\n",
    "plt.xscale('log')\n",
    "plt.yscale('log')\n",
    "plt.xlabel('d/n')\n",
    "plt.tight_layout()\n",
    "plt.savefig('mdl_theory.pdf')\n",
    "plt.show()\n",
    "\n",
    "plt.figure(figsize=[8, 6])\n",
    "for j, d in enumerate(ds):\n",
    "    ns = np.asarray(d/gammas, dtype=int)\n",
    "    plt.plot(d/ns, mdl_practice.mean(0)[j, :], color=cs[j], linestyle=lss[j],linewidth=4.,  label='d=%d'%(d))\n",
    "plt.title('Prac-MDL-COMP', fontweight='bold', fontsize=20)\n",
    "plt.xscale('log')\n",
    "plt.legend(frameon=False)\n",
    "plt.xlabel('d/n')\n",
    "plt.grid(True, alpha=0.5, linewidth=1.)\n",
    "plt.tight_layout()\n",
    "\n",
    "plt.savefig('mdl_practical.pdf')\n",
    "plt.show()\n",
    "\n",
    "\n",
    "plt.figure(figsize=[8, 6])\n",
    "d = ds[0]\n",
    "ns = np.asarray(d/gammas, dtype=int)\n",
    "plt.plot(d/ns, mdl_theory.mean(0)[0, :], color=cs[0], linestyle='-',linewidth=4.,  label='MDL-COMP')\n",
    "plt.plot(d/ns, thm3_bnds[d], color=cs[0], linestyle=':',linewidth=4.,  label='(RMT) MDL-COMP')\n",
    "plt.plot(d/ns, mdl_practice.mean(0)[0, :], color=cs[0], linestyle='--',linewidth=4.,  label='Prac-MDL-COMP')\n",
    "plt.legend(frameon=False)\n",
    "plt.title('Comparison of MDL-complexities (d=%d)'%d, fontweight='bold', fontsize=20)\n",
    "plt.xscale('log')\n",
    "plt.yscale('log')\n",
    "plt.xlabel('d/n')\n",
    "plt.grid(True, alpha=0.5, linewidth=1.)\n",
    "plt.tight_layout()\n",
    "\n",
    "plt.savefig('mdl_comparison_all_bounds.pdf')\n",
    "plt.show()\n",
    "\n",
    "\n",
    "\n",
    "plt.figure(figsize=[8, 6])\n",
    "plt.plot(d/ns, mdl_practice.mean(0)[0, :], color=cs[0], linestyle='-', linewidth=4.,  label='Prac-MDL-COMP')\n",
    "plt.plot(d/ns, mse.mean(0)[0, :], color=cs[1], linestyle='-.', linewidth=4.,label='In-sample MSE')\n",
    "plt.plot(d/ns, eigsum.mean(0)[0, :], color=cs[2], linestyle='--',linewidth=4.,  label='Eig. Sum')\n",
    "plt.plot(d/ns, tnorm.mean(0)[0, :], color=cs[3], linestyle=':',  linewidth=4., label='Theta Norm')\n",
    "plt.xscale('log')\n",
    "plt.xlabel('d/n')\n",
    "# plt.ylim([0., 1.75])\n",
    "plt.grid(True, alpha=0.5, linewidth=1.)\n",
    "plt.title('Decomposing Prac-MDL-COMP (d=%d)'%d, fontweight='bold', fontsize=20)\n",
    "plt.legend(frameon=False)\n",
    "plt.tight_layout()\n",
    "plt.savefig('mdl_practical_details.pdf')\n",
    "plt.show()\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Run simulations"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "n = 400"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "  0%|          | 0/10 [00:00<?, ?it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "100 2000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 10%|█         | 1/10 [00:00<00:07,  1.17it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 20%|██        | 2/10 [00:01<00:06,  1.19it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 30%|███       | 3/10 [00:02<00:05,  1.21it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 40%|████      | 4/10 [00:03<00:04,  1.22it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 50%|█████     | 5/10 [00:04<00:04,  1.23it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 60%|██████    | 6/10 [00:04<00:03,  1.24it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 70%|███████   | 7/10 [00:05<00:02,  1.24it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 80%|████████  | 8/10 [00:06<00:01,  1.25it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 90%|█████████ | 9/10 [00:07<00:00,  1.26it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "100%|██████████| 10/10 [00:08<00:00,  1.26it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "  0%|          | 0/10 [00:00<?, ?it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "100 1000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 10%|█         | 1/10 [00:00<00:05,  1.64it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 20%|██        | 2/10 [00:01<00:04,  1.63it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 30%|███       | 3/10 [00:01<00:04,  1.63it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 40%|████      | 4/10 [00:02<00:03,  1.63it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 50%|█████     | 5/10 [00:03<00:03,  1.63it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 60%|██████    | 6/10 [00:03<00:02,  1.65it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 70%|███████   | 7/10 [00:04<00:01,  1.64it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 80%|████████  | 8/10 [00:04<00:01,  1.64it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 90%|█████████ | 9/10 [00:05<00:00,  1.64it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "100%|██████████| 10/10 [00:06<00:00,  1.65it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "  0%|          | 0/10 [00:00<?, ?it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "100 500\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 10%|█         | 1/10 [00:00<00:04,  1.93it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 20%|██        | 2/10 [00:01<00:04,  1.93it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 30%|███       | 3/10 [00:01<00:03,  1.94it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 40%|████      | 4/10 [00:02<00:03,  1.93it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 50%|█████     | 5/10 [00:02<00:02,  1.92it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 60%|██████    | 6/10 [00:03<00:02,  1.93it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 70%|███████   | 7/10 [00:03<00:01,  1.94it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 80%|████████  | 8/10 [00:04<00:01,  1.95it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 90%|█████████ | 9/10 [00:04<00:00,  1.96it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "100%|██████████| 10/10 [00:05<00:00,  1.96it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "  0%|          | 0/10 [00:00<?, ?it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "100 200\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 10%|█         | 1/10 [00:00<00:03,  2.27it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 20%|██        | 2/10 [00:00<00:03,  2.27it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 30%|███       | 3/10 [00:01<00:03,  2.24it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 40%|████      | 4/10 [00:01<00:02,  2.23it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 50%|█████     | 5/10 [00:02<00:02,  2.24it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 60%|██████    | 6/10 [00:02<00:01,  2.24it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 70%|███████   | 7/10 [00:03<00:01,  2.23it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 80%|████████  | 8/10 [00:03<00:00,  2.22it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 90%|█████████ | 9/10 [00:04<00:00,  2.22it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "100%|██████████| 10/10 [00:04<00:00,  2.20it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "  0%|          | 0/10 [00:00<?, ?it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "100 133\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 10%|█         | 1/10 [00:00<00:03,  2.40it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 20%|██        | 2/10 [00:00<00:03,  2.38it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 30%|███       | 3/10 [00:01<00:02,  2.36it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 40%|████      | 4/10 [00:01<00:02,  2.37it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 50%|█████     | 5/10 [00:02<00:02,  2.34it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 60%|██████    | 6/10 [00:02<00:01,  2.33it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 70%|███████   | 7/10 [00:02<00:01,  2.33it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 80%|████████  | 8/10 [00:03<00:00,  2.33it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 90%|█████████ | 9/10 [00:03<00:00,  2.32it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "100%|██████████| 10/10 [00:04<00:00,  2.32it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "  0%|          | 0/10 [00:00<?, ?it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "100 111\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 10%|█         | 1/10 [00:00<00:03,  2.56it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 20%|██        | 2/10 [00:00<00:03,  2.53it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 30%|███       | 3/10 [00:01<00:02,  2.53it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 40%|████      | 4/10 [00:01<00:02,  2.52it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 50%|█████     | 5/10 [00:01<00:01,  2.50it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 60%|██████    | 6/10 [00:02<00:01,  2.47it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 70%|███████   | 7/10 [00:02<00:01,  2.47it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 80%|████████  | 8/10 [00:03<00:00,  2.45it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 90%|█████████ | 9/10 [00:03<00:00,  2.43it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "100%|██████████| 10/10 [00:04<00:00,  2.43it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "  0%|          | 0/10 [00:00<?, ?it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "100 105\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 10%|█         | 1/10 [00:00<00:03,  2.51it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 20%|██        | 2/10 [00:00<00:03,  2.47it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 30%|███       | 3/10 [00:01<00:02,  2.47it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 40%|████      | 4/10 [00:01<00:02,  2.46it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 50%|█████     | 5/10 [00:02<00:02,  2.45it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 60%|██████    | 6/10 [00:02<00:01,  2.44it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 70%|███████   | 7/10 [00:02<00:01,  2.44it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 80%|████████  | 8/10 [00:03<00:00,  2.45it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 90%|█████████ | 9/10 [00:03<00:00,  2.46it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "100%|██████████| 10/10 [00:04<00:00,  2.45it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "  0%|          | 0/10 [00:00<?, ?it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "100 100\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 10%|█         | 1/10 [00:00<00:03,  2.83it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 20%|██        | 2/10 [00:00<00:02,  2.74it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 30%|███       | 3/10 [00:01<00:02,  2.85it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 40%|████      | 4/10 [00:01<00:02,  2.80it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 50%|█████     | 5/10 [00:01<00:01,  2.80it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 60%|██████    | 6/10 [00:02<00:01,  2.83it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 70%|███████   | 7/10 [00:02<00:01,  2.80it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 80%|████████  | 8/10 [00:02<00:00,  2.78it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 90%|█████████ | 9/10 [00:03<00:00,  2.83it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "100%|██████████| 10/10 [00:03<00:00,  2.85it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "  0%|          | 0/10 [00:00<?, ?it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "100 95\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/linux/anaconda3.7/lib/python3.7/site-packages/ipykernel_launcher.py:46: ComplexWarning: Casting complex values to real discards the imaginary part\n",
      "/usr/local/linux/anaconda3.7/lib/python3.7/site-packages/ipykernel_launcher.py:47: ComplexWarning: Casting complex values to real discards the imaginary part\n",
      "/usr/local/linux/anaconda3.7/lib/python3.7/site-packages/ipykernel_launcher.py:48: ComplexWarning: Casting complex values to real discards the imaginary part\n",
      "/usr/local/linux/anaconda3.7/lib/python3.7/site-packages/ipykernel_launcher.py:49: ComplexWarning: Casting complex values to real discards the imaginary part\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 10%|█         | 1/10 [00:00<00:03,  2.53it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 20%|██        | 2/10 [00:00<00:03,  2.51it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 30%|███       | 3/10 [00:01<00:02,  2.51it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 40%|████      | 4/10 [00:01<00:02,  2.60it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 50%|█████     | 5/10 [00:01<00:01,  2.59it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 60%|██████    | 6/10 [00:02<00:01,  2.57it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 70%|███████   | 7/10 [00:02<00:01,  2.57it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 80%|████████  | 8/10 [00:03<00:00,  2.55it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 90%|█████████ | 9/10 [00:03<00:00,  2.56it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "100%|██████████| 10/10 [00:03<00:00,  2.54it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "  0%|          | 0/10 [00:00<?, ?it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "100 80\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 10%|█         | 1/10 [00:00<00:03,  2.87it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 20%|██        | 2/10 [00:00<00:02,  2.84it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 30%|███       | 3/10 [00:01<00:02,  2.82it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 40%|████      | 4/10 [00:01<00:02,  2.83it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 50%|█████     | 5/10 [00:01<00:01,  2.81it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 60%|██████    | 6/10 [00:02<00:01,  2.80it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 70%|███████   | 7/10 [00:02<00:01,  2.79it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 80%|████████  | 8/10 [00:02<00:00,  2.78it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 90%|█████████ | 9/10 [00:03<00:00,  2.78it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "100%|██████████| 10/10 [00:03<00:00,  2.79it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "  0%|          | 0/10 [00:00<?, ?it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "100 66\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 10%|█         | 1/10 [00:00<00:02,  3.07it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 20%|██        | 2/10 [00:00<00:02,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 30%|███       | 3/10 [00:00<00:02,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 40%|████      | 4/10 [00:01<00:01,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 50%|█████     | 5/10 [00:01<00:01,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 60%|██████    | 6/10 [00:01<00:01,  3.07it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 70%|███████   | 7/10 [00:02<00:00,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 80%|████████  | 8/10 [00:02<00:00,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 90%|█████████ | 9/10 [00:02<00:00,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "100%|██████████| 10/10 [00:03<00:00,  3.06it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "  0%|          | 0/10 [00:00<?, ?it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "100 57\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 10%|█         | 1/10 [00:00<00:02,  3.38it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 20%|██        | 2/10 [00:00<00:02,  3.34it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 30%|███       | 3/10 [00:00<00:02,  3.34it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 40%|████      | 4/10 [00:01<00:01,  3.28it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 50%|█████     | 5/10 [00:01<00:01,  3.29it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 60%|██████    | 6/10 [00:01<00:01,  3.30it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 70%|███████   | 7/10 [00:02<00:00,  3.03it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 80%|████████  | 8/10 [00:02<00:00,  3.07it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 90%|█████████ | 9/10 [00:02<00:00,  3.15it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "100%|██████████| 10/10 [00:03<00:00,  3.20it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "  0%|          | 0/10 [00:00<?, ?it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "100 51\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 10%|█         | 1/10 [00:00<00:02,  3.48it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 20%|██        | 2/10 [00:00<00:02,  3.44it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 30%|███       | 3/10 [00:00<00:02,  3.42it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 40%|████      | 4/10 [00:01<00:01,  3.42it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 50%|█████     | 5/10 [00:01<00:01,  3.40it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 60%|██████    | 6/10 [00:01<00:01,  3.41it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 70%|███████   | 7/10 [00:02<00:00,  3.41it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 80%|████████  | 8/10 [00:02<00:00,  3.40it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 90%|█████████ | 9/10 [00:02<00:00,  3.39it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "100%|██████████| 10/10 [00:02<00:00,  3.40it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "  0%|          | 0/10 [00:00<?, ?it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "100 50\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 10%|█         | 1/10 [00:00<00:02,  3.45it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 20%|██        | 2/10 [00:00<00:02,  3.43it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 30%|███       | 3/10 [00:00<00:02,  3.43it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 40%|████      | 4/10 [00:01<00:01,  3.41it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 50%|█████     | 5/10 [00:01<00:01,  3.40it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 60%|██████    | 6/10 [00:01<00:01,  3.42it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 70%|███████   | 7/10 [00:02<00:00,  3.44it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 80%|████████  | 8/10 [00:02<00:00,  3.45it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 90%|█████████ | 9/10 [00:02<00:00,  3.43it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "100%|██████████| 10/10 [00:02<00:00,  3.44it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "  0%|          | 0/10 [00:00<?, ?it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "100 48\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 10%|█         | 1/10 [00:00<00:02,  3.54it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 20%|██        | 2/10 [00:00<00:02,  3.52it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 30%|███       | 3/10 [00:00<00:01,  3.51it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 40%|████      | 4/10 [00:01<00:01,  3.50it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 50%|█████     | 5/10 [00:01<00:01,  3.48it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 60%|██████    | 6/10 [00:01<00:01,  3.48it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 70%|███████   | 7/10 [00:02<00:00,  3.48it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 80%|████████  | 8/10 [00:02<00:00,  3.51it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 90%|█████████ | 9/10 [00:02<00:00,  3.50it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "100%|██████████| 10/10 [00:02<00:00,  3.50it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "  0%|          | 0/10 [00:00<?, ?it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "100 44\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 10%|█         | 1/10 [00:00<00:02,  3.58it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 20%|██        | 2/10 [00:00<00:02,  3.55it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 30%|███       | 3/10 [00:00<00:01,  3.53it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 40%|████      | 4/10 [00:01<00:01,  3.53it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 50%|█████     | 5/10 [00:01<00:01,  3.52it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 60%|██████    | 6/10 [00:01<00:01,  3.50it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 70%|███████   | 7/10 [00:01<00:00,  3.51it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 80%|████████  | 8/10 [00:02<00:00,  3.51it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 90%|█████████ | 9/10 [00:02<00:00,  3.49it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "100%|██████████| 10/10 [00:02<00:00,  3.48it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "  0%|          | 0/10 [00:00<?, ?it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "100 40\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 10%|█         | 1/10 [00:00<00:02,  3.72it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 20%|██        | 2/10 [00:00<00:02,  3.68it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 30%|███       | 3/10 [00:00<00:01,  3.68it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 40%|████      | 4/10 [00:01<00:01,  3.67it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 50%|█████     | 5/10 [00:01<00:01,  3.68it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 60%|██████    | 6/10 [00:01<00:01,  3.68it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 70%|███████   | 7/10 [00:01<00:00,  3.68it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 80%|████████  | 8/10 [00:02<00:00,  3.67it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 90%|█████████ | 9/10 [00:02<00:00,  3.67it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "100%|██████████| 10/10 [00:02<00:00,  3.67it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "  0%|          | 0/10 [00:00<?, ?it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "100 20\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 10%|█         | 1/10 [00:00<00:02,  4.44it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 20%|██        | 2/10 [00:00<00:01,  4.44it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 30%|███       | 3/10 [00:00<00:01,  4.41it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 40%|████      | 4/10 [00:00<00:01,  4.42it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 50%|█████     | 5/10 [00:01<00:01,  4.37it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 60%|██████    | 6/10 [00:01<00:00,  4.34it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 70%|███████   | 7/10 [00:01<00:00,  4.38it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 80%|████████  | 8/10 [00:01<00:00,  4.38it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 90%|█████████ | 9/10 [00:02<00:00,  4.37it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "100%|██████████| 10/10 [00:02<00:00,  4.37it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "  0%|          | 0/10 [00:00<?, ?it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "100 10\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 10%|█         | 1/10 [00:00<00:01,  4.99it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 20%|██        | 2/10 [00:00<00:01,  4.91it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 30%|███       | 3/10 [00:00<00:01,  4.92it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 40%|████      | 4/10 [00:00<00:01,  4.88it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 50%|█████     | 5/10 [00:01<00:01,  4.90it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 60%|██████    | 6/10 [00:01<00:00,  4.94it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 70%|███████   | 7/10 [00:01<00:00,  4.98it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 80%|████████  | 8/10 [00:01<00:00,  4.99it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 90%|█████████ | 9/10 [00:01<00:00,  5.00it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "100%|██████████| 10/10 [00:02<00:00,  4.94it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "  0%|          | 0/10 [00:00<?, ?it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "100 5\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 10%|█         | 1/10 [00:00<00:01,  5.01it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 20%|██        | 2/10 [00:00<00:01,  5.06it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 30%|███       | 3/10 [00:00<00:01,  4.84it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 40%|████      | 4/10 [00:00<00:01,  4.93it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 50%|█████     | 5/10 [00:01<00:00,  5.03it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 60%|██████    | 6/10 [00:01<00:00,  5.03it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 70%|███████   | 7/10 [00:01<00:00,  5.04it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 80%|████████  | 8/10 [00:01<00:00,  5.07it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 90%|█████████ | 9/10 [00:01<00:00,  5.03it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "100%|██████████| 10/10 [00:01<00:00,  5.03it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "  0%|          | 0/10 [00:00<?, ?it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "200 4000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 10%|█         | 1/10 [00:04<00:39,  4.34s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 20%|██        | 2/10 [00:08<00:34,  4.34s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 30%|███       | 3/10 [00:12<00:30,  4.30s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 40%|████      | 4/10 [00:17<00:25,  4.27s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 50%|█████     | 5/10 [00:21<00:21,  4.29s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 60%|██████    | 6/10 [00:25<00:17,  4.27s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 70%|███████   | 7/10 [00:29<00:12,  4.29s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 80%|████████  | 8/10 [00:34<00:08,  4.29s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 90%|█████████ | 9/10 [00:38<00:04,  4.30s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "100%|██████████| 10/10 [00:42<00:00,  4.30s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "  0%|          | 0/10 [00:00<?, ?it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "200 2000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 10%|█         | 1/10 [00:02<00:26,  2.92s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 20%|██        | 2/10 [00:05<00:23,  2.95s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 30%|███       | 3/10 [00:08<00:20,  2.96s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 40%|████      | 4/10 [00:11<00:17,  2.97s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 50%|█████     | 5/10 [00:14<00:14,  2.98s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 60%|██████    | 6/10 [00:17<00:11,  2.99s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 70%|███████   | 7/10 [00:20<00:08,  2.99s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 80%|████████  | 8/10 [00:23<00:05,  2.99s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 90%|█████████ | 9/10 [00:26<00:02,  2.99s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "100%|██████████| 10/10 [00:29<00:00,  2.99s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "  0%|          | 0/10 [00:00<?, ?it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "200 1000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 10%|█         | 1/10 [00:02<00:20,  2.32s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 20%|██        | 2/10 [00:04<00:18,  2.32s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 30%|███       | 3/10 [00:06<00:16,  2.32s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 40%|████      | 4/10 [00:09<00:13,  2.32s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 50%|█████     | 5/10 [00:11<00:11,  2.32s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 60%|██████    | 6/10 [00:13<00:09,  2.32s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 70%|███████   | 7/10 [00:16<00:06,  2.32s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 80%|████████  | 8/10 [00:18<00:04,  2.32s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 90%|█████████ | 9/10 [00:20<00:02,  2.32s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "100%|██████████| 10/10 [00:23<00:00,  2.34s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "  0%|          | 0/10 [00:00<?, ?it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "200 400\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 10%|█         | 1/10 [00:01<00:16,  1.89s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 20%|██        | 2/10 [00:03<00:15,  1.94s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 30%|███       | 3/10 [00:05<00:13,  1.94s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 40%|████      | 4/10 [00:07<00:11,  1.92s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 50%|█████     | 5/10 [00:09<00:09,  1.91s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 60%|██████    | 6/10 [00:11<00:07,  1.92s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 70%|███████   | 7/10 [00:13<00:05,  1.91s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 80%|████████  | 8/10 [00:15<00:03,  1.92s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 90%|█████████ | 9/10 [00:17<00:01,  1.91s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "100%|██████████| 10/10 [00:19<00:00,  1.92s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "  0%|          | 0/10 [00:00<?, ?it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "200 266\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 10%|█         | 1/10 [00:01<00:16,  1.81s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 20%|██        | 2/10 [00:03<00:14,  1.82s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 30%|███       | 3/10 [00:05<00:12,  1.82s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 40%|████      | 4/10 [00:07<00:10,  1.82s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 50%|█████     | 5/10 [00:09<00:09,  1.82s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 60%|██████    | 6/10 [00:10<00:07,  1.82s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 70%|███████   | 7/10 [00:12<00:05,  1.82s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 80%|████████  | 8/10 [00:14<00:03,  1.81s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 90%|█████████ | 9/10 [00:16<00:01,  1.81s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "100%|██████████| 10/10 [00:18<00:00,  1.81s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "  0%|          | 0/10 [00:00<?, ?it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "200 222\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 10%|█         | 1/10 [00:01<00:15,  1.73s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 20%|██        | 2/10 [00:03<00:13,  1.72s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 30%|███       | 3/10 [00:05<00:11,  1.71s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 40%|████      | 4/10 [00:06<00:10,  1.71s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 50%|█████     | 5/10 [00:08<00:08,  1.70s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 60%|██████    | 6/10 [00:10<00:06,  1.71s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 70%|███████   | 7/10 [00:11<00:05,  1.70s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 80%|████████  | 8/10 [00:13<00:03,  1.71s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 90%|█████████ | 9/10 [00:15<00:01,  1.71s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "100%|██████████| 10/10 [00:17<00:00,  1.70s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "  0%|          | 0/10 [00:00<?, ?it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "200 210\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 10%|█         | 1/10 [00:01<00:15,  1.71s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 20%|██        | 2/10 [00:03<00:13,  1.71s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 30%|███       | 3/10 [00:05<00:12,  1.71s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 40%|████      | 4/10 [00:06<00:10,  1.71s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 50%|█████     | 5/10 [00:08<00:08,  1.70s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 60%|██████    | 6/10 [00:10<00:06,  1.70s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 70%|███████   | 7/10 [00:11<00:05,  1.69s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 80%|████████  | 8/10 [00:13<00:03,  1.69s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 90%|█████████ | 9/10 [00:15<00:01,  1.69s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "100%|██████████| 10/10 [00:17<00:00,  1.70s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "  0%|          | 0/10 [00:00<?, ?it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "200 200\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 10%|█         | 1/10 [00:01<00:13,  1.51s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 20%|██        | 2/10 [00:03<00:12,  1.54s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 30%|███       | 3/10 [00:04<00:10,  1.55s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 40%|████      | 4/10 [00:06<00:09,  1.54s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 50%|█████     | 5/10 [00:07<00:07,  1.54s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 60%|██████    | 6/10 [00:09<00:06,  1.57s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 70%|███████   | 7/10 [00:10<00:04,  1.57s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 80%|████████  | 8/10 [00:12<00:03,  1.56s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 90%|█████████ | 9/10 [00:14<00:01,  1.60s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "100%|██████████| 10/10 [00:15<00:00,  1.61s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "  0%|          | 0/10 [00:00<?, ?it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "200 190\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 10%|█         | 1/10 [00:01<00:15,  1.69s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 20%|██        | 2/10 [00:03<00:13,  1.68s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 30%|███       | 3/10 [00:04<00:11,  1.67s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 40%|████      | 4/10 [00:06<00:09,  1.65s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 50%|█████     | 5/10 [00:08<00:08,  1.65s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 60%|██████    | 6/10 [00:09<00:06,  1.64s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 70%|███████   | 7/10 [00:11<00:04,  1.64s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 80%|████████  | 8/10 [00:13<00:03,  1.66s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 90%|█████████ | 9/10 [00:14<00:01,  1.65s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "100%|██████████| 10/10 [00:16<00:00,  1.64s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "  0%|          | 0/10 [00:00<?, ?it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "200 160\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 10%|█         | 1/10 [00:01<00:13,  1.48s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 20%|██        | 2/10 [00:02<00:11,  1.49s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 30%|███       | 3/10 [00:04<00:10,  1.50s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 40%|████      | 4/10 [00:06<00:09,  1.50s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 50%|█████     | 5/10 [00:07<00:07,  1.51s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 60%|██████    | 6/10 [00:09<00:06,  1.51s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 70%|███████   | 7/10 [00:10<00:04,  1.51s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 80%|████████  | 8/10 [00:12<00:03,  1.52s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 90%|█████████ | 9/10 [00:13<00:01,  1.52s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "100%|██████████| 10/10 [00:15<00:00,  1.52s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "  0%|          | 0/10 [00:00<?, ?it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "200 133\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 10%|█         | 1/10 [00:01<00:12,  1.40s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 20%|██        | 2/10 [00:02<00:11,  1.40s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 30%|███       | 3/10 [00:04<00:09,  1.41s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 40%|████      | 4/10 [00:05<00:08,  1.43s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 50%|█████     | 5/10 [00:07<00:07,  1.42s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 60%|██████    | 6/10 [00:08<00:05,  1.41s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 70%|███████   | 7/10 [00:09<00:04,  1.41s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 80%|████████  | 8/10 [00:11<00:02,  1.41s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 90%|█████████ | 9/10 [00:12<00:01,  1.42s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "100%|██████████| 10/10 [00:14<00:00,  1.41s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "  0%|          | 0/10 [00:00<?, ?it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "200 114\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 10%|█         | 1/10 [00:01<00:12,  1.39s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 20%|██        | 2/10 [00:02<00:11,  1.40s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 30%|███       | 3/10 [00:04<00:09,  1.39s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 40%|████      | 4/10 [00:05<00:08,  1.39s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 50%|█████     | 5/10 [00:06<00:06,  1.39s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 60%|██████    | 6/10 [00:08<00:05,  1.39s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 70%|███████   | 7/10 [00:09<00:04,  1.39s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 80%|████████  | 8/10 [00:11<00:02,  1.38s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 90%|█████████ | 9/10 [00:12<00:01,  1.38s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "100%|██████████| 10/10 [00:13<00:00,  1.37s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "  0%|          | 0/10 [00:00<?, ?it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "200 102\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 10%|█         | 1/10 [00:01<00:11,  1.31s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 20%|██        | 2/10 [00:02<00:10,  1.31s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 30%|███       | 3/10 [00:03<00:09,  1.31s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 40%|████      | 4/10 [00:05<00:07,  1.31s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 50%|█████     | 5/10 [00:06<00:06,  1.32s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 60%|██████    | 6/10 [00:07<00:05,  1.32s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 70%|███████   | 7/10 [00:09<00:03,  1.32s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 80%|████████  | 8/10 [00:10<00:02,  1.32s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 90%|█████████ | 9/10 [00:11<00:01,  1.32s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "100%|██████████| 10/10 [00:13<00:00,  1.32s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "  0%|          | 0/10 [00:00<?, ?it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "200 100\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 10%|█         | 1/10 [00:01<00:11,  1.29s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 20%|██        | 2/10 [00:02<00:10,  1.30s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 30%|███       | 3/10 [00:03<00:09,  1.30s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 40%|████      | 4/10 [00:05<00:07,  1.30s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 50%|█████     | 5/10 [00:06<00:06,  1.30s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 60%|██████    | 6/10 [00:07<00:05,  1.31s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 70%|███████   | 7/10 [00:09<00:03,  1.31s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 80%|████████  | 8/10 [00:10<00:02,  1.30s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 90%|█████████ | 9/10 [00:11<00:01,  1.32s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "100%|██████████| 10/10 [00:13<00:00,  1.32s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "  0%|          | 0/10 [00:00<?, ?it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "200 97\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 10%|█         | 1/10 [00:01<00:11,  1.30s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 20%|██        | 2/10 [00:02<00:10,  1.30s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 30%|███       | 3/10 [00:03<00:09,  1.30s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 40%|████      | 4/10 [00:05<00:07,  1.30s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 50%|█████     | 5/10 [00:06<00:06,  1.29s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 60%|██████    | 6/10 [00:07<00:05,  1.28s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 70%|███████   | 7/10 [00:08<00:03,  1.28s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 80%|████████  | 8/10 [00:10<00:02,  1.27s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 90%|█████████ | 9/10 [00:11<00:01,  1.27s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "100%|██████████| 10/10 [00:12<00:00,  1.27s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "  0%|          | 0/10 [00:00<?, ?it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "200 88\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 10%|█         | 1/10 [00:01<00:10,  1.22s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 20%|██        | 2/10 [00:02<00:09,  1.22s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 30%|███       | 3/10 [00:03<00:08,  1.22s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 40%|████      | 4/10 [00:04<00:07,  1.23s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 50%|█████     | 5/10 [00:06<00:06,  1.23s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 60%|██████    | 6/10 [00:07<00:04,  1.23s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 70%|███████   | 7/10 [00:08<00:03,  1.24s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 80%|████████  | 8/10 [00:09<00:02,  1.24s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 90%|█████████ | 9/10 [00:11<00:01,  1.24s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "100%|██████████| 10/10 [00:12<00:00,  1.24s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "  0%|          | 0/10 [00:00<?, ?it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "200 80\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 10%|█         | 1/10 [00:01<00:11,  1.23s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 20%|██        | 2/10 [00:02<00:09,  1.22s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 30%|███       | 3/10 [00:03<00:08,  1.23s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 40%|████      | 4/10 [00:04<00:07,  1.22s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 50%|█████     | 5/10 [00:06<00:06,  1.22s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 60%|██████    | 6/10 [00:07<00:04,  1.22s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 70%|███████   | 7/10 [00:08<00:03,  1.22s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 80%|████████  | 8/10 [00:09<00:02,  1.23s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 90%|█████████ | 9/10 [00:10<00:01,  1.22s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "100%|██████████| 10/10 [00:12<00:00,  1.22s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "  0%|          | 0/10 [00:00<?, ?it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "200 40\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 10%|█         | 1/10 [00:01<00:09,  1.06s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 20%|██        | 2/10 [00:02<00:08,  1.06s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 30%|███       | 3/10 [00:03<00:07,  1.06s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 40%|████      | 4/10 [00:04<00:06,  1.06s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 50%|█████     | 5/10 [00:05<00:05,  1.06s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 60%|██████    | 6/10 [00:06<00:04,  1.06s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 70%|███████   | 7/10 [00:07<00:03,  1.07s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 80%|████████  | 8/10 [00:08<00:02,  1.07s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 90%|█████████ | 9/10 [00:09<00:01,  1.06s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "100%|██████████| 10/10 [00:10<00:00,  1.07s/it]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "  0%|          | 0/10 [00:00<?, ?it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "200 20\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 10%|█         | 1/10 [00:00<00:08,  1.02it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 20%|██        | 2/10 [00:01<00:07,  1.02it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 30%|███       | 3/10 [00:02<00:06,  1.02it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 40%|████      | 4/10 [00:03<00:05,  1.02it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 50%|█████     | 5/10 [00:04<00:04,  1.02it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 60%|██████    | 6/10 [00:05<00:03,  1.01it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 70%|███████   | 7/10 [00:06<00:02,  1.01it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 80%|████████  | 8/10 [00:07<00:01,  1.01it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 90%|█████████ | 9/10 [00:08<00:00,  1.02it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "100%|██████████| 10/10 [00:09<00:00,  1.03it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "  0%|          | 0/10 [00:00<?, ?it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "200 10\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 10%|█         | 1/10 [00:00<00:08,  1.06it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 20%|██        | 2/10 [00:01<00:07,  1.05it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 30%|███       | 3/10 [00:02<00:06,  1.06it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 40%|████      | 4/10 [00:03<00:05,  1.05it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 50%|█████     | 5/10 [00:04<00:04,  1.05it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 60%|██████    | 6/10 [00:05<00:03,  1.05it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 70%|███████   | 7/10 [00:06<00:02,  1.04it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 80%|████████  | 8/10 [00:07<00:01,  1.05it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " 90%|█████████ | 9/10 [00:08<00:00,  1.06it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "100%|██████████| 10/10 [00:09<00:00,  1.05it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    }
   ],
   "source": [
    "num_trials = 10\n",
    "gamma_inv = np.array([5e-2, 1e-1, 0.5, 0.75, 0.9, 1,\n",
    "                                  1.2, 1.5, 2, 5, 7.5, 1e1, 2e1, 4e1])\n",
    "\n",
    "gammas = np.array([5e-2, 1e-1, 0.2, 0.5, 0.75, 0.9, 0.95, 1.,  1.05,  1.25, 1.5,  1.75, 1.95, 2., 2.05, 2.25, 2.5, 5., 10., 20.])\n",
    "\n",
    "noise_std = 0.1\n",
    "n_test = 1000\n",
    "\n",
    "ds = [100, 200]\n",
    "ns = np.asarray(ds[0]/gammas, dtype=int)\n",
    "\n",
    "mdl_practice = np.zeros((num_trials,  len(ds), len(ns)))\n",
    "mdl_theory = np.zeros((num_trials, len(ds), len(ns)))\n",
    "mdl_theory_mean = np.zeros((num_trials, len(ds), len(ns)))\n",
    "mdl_theory_lam  = np.zeros((num_trials, len(ds), len(ns)))\n",
    "mse = np.zeros((num_trials, len(ds), len(ns)))\n",
    "tnorm = np.zeros((num_trials, len(ds), len(ns)))\n",
    "eigsum = np.zeros((num_trials, len(ds), len(ns)))\n",
    "mse_test = np.zeros((num_trials, len(ds), len(ns)))\n",
    "\n",
    "for j, d in enumerate(ds):\n",
    "    ns = np.asarray(d/gammas, dtype=int)\n",
    "    for k, n in enumerate(ns):\n",
    "        print(d, n)\n",
    "        X_train = npr.randn(n, d) / np.sqrt(n)\n",
    "        X_test = npr.randn(n_test, d)\n",
    "        eigenvals, eigenvecs = npl.eig(X_train.T @ X_train)\n",
    "        \n",
    "        for i in tqdm(range(num_trials)):\n",
    "            \n",
    "            beta_star = npr.randn(d)\n",
    "            beta_star /= npl.norm(beta_star)    \n",
    "            beta_star *= 1. #\n",
    "\n",
    "            w = eigenvecs.T  @ beta_star\n",
    "            \n",
    "            y_train = X_train @ beta_star + noise_std * npr.randn(n)\n",
    "            y_test =  X_test @ beta_star # + noise_std *  npr.randn(n_test)\n",
    "\n",
    "            var = noise_std**2\n",
    "            opt_solved = minimize(loc_mdl1_loss, x0=1e-10)\n",
    "            lambda_opt = opt_solved.x\n",
    "            \n",
    "            mdl_practice[i, j, k] = opt_solved.fun\n",
    "            mdl_theory[i, j, k] = np.sum(np.log(1+ eigenvals * w**2 / var)) / n\n",
    "            mdl_theory_mean[i, j, k] = np.sum(np.log(1+ eigenvals * np.mean(w**2) /var)) / n\n",
    "            mdl_theory_lam[i, j, k] = np.sum(np.log(1+ eigenvals / lambda_opt)) / n\n",
    "            mse[i, j, k], tnorm[i, j, k], eigsum[i, j, k] = loc_mdl1_loss_ret_all(lambda_opt)\n",
    "            \n",
    "            mse_test[i, j, k] = npl.norm(y_test- X_test @ loc_calc_thetahat(lambda_opt))**2 / n_test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "metadata": {},
   "outputs": [],
   "source": [
    "thm3_bnds = dict()\n",
    "for d in ds:\n",
    "    ns = np.asarray(d/gammas, dtype=int)\n",
    "    thm3_bnds[d] = thm3(d/ns, npl.norm(beta_star)/var/d)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1296x288 with 4 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=[18, 4])\n",
    "\n",
    "for j, d in enumerate(ds):\n",
    "    ns = np.asarray(d/gammas, dtype=int)\n",
    "    plt.subplot(1, 4, j+1)\n",
    "    plt.plot(d/ns, mdl_practice.mean(0)[j, :], color=cs[0], linestyle='-', linewidth=3., label='Practical Bound')\n",
    "    plt.plot(d/ns, mdl_theory.mean(0)[j, :], color=cs[0], linestyle='--',linewidth=3.,  label='Thm-1 Bound')\n",
    "    plt.plot(d/ns, thm3_bnds[d], color=cs[0], linestyle=':',linewidth=3.,  label='Thm-3 Bound')\n",
    "    plt.xscale('log')\n",
    "    plt.yscale('log')\n",
    "    plt.xlabel('d/n')\n",
    "    if j==0:\n",
    "        plt.legend()\n",
    "    \n",
    "    plt.subplot(1, 4, j+3)\n",
    "    plt.plot(d/ns, mdl_practice.mean(0)[j, :], color=cs[j], linestyle='-', marker='D', label='MDL-COMP-Practical')\n",
    "    plt.plot(d/ns, mse.mean(0)[j, :], color=cs[1], linestyle='-.', marker='D', label='mse')\n",
    "    plt.plot(d/ns, eigsum.mean(0)[j, :], color=cs[2], linestyle='--',  marker='o', label='eigsum')\n",
    "    plt.plot(d/ns, tnorm.mean(0)[j, :], color=cs[3], linestyle=':',  marker='s', label='tnorm')\n",
    "#     plt.plot(d/ns, tnorm.mean(0)[j, :]+mse.mean(0)[j, :]+eigsum.mean(0)[j, :], 'k--')\n",
    "    plt.xscale('log')\n",
    "    plt.xlabel('d/n')\n",
    "plt.tight_layout()\n",
    "plt.show()\n",
    "#     plt.subplot(122)\n",
    "#     plt.loglog(d/ns, mdl_practice.mean(0)[j, :]/mdl_theory.mean(0)[j, :], color=cs[j])\n",
    "# plt.xscale('log')\n",
    "#     plt.legend()\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Change d"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "n=400\n",
      "d=40\n",
      "d=67\n",
      "d=112\n",
      "d=189\n",
      "d=317\n",
      "d=356\n",
      "d=400\n",
      "d=448\n",
      "d=503\n",
      "d=633\n",
      "d=1004\n",
      "d=1592\n",
      "d=2523\n",
      "d=4000\n"
     ]
    }
   ],
   "source": [
    "npr.seed(15430)\n",
    "num_trials = 5\n",
    "\n",
    "gammas = np.hstack((np.logspace(-1, -0.1, 5),\n",
    "        np.logspace(-0.05, 0.1, 4),\n",
    "        np.logspace(0.2, 1., 5)))\n",
    "\n",
    "# gammas = np.array([0.1, 0.5, 1., 2., 5.])\n",
    "noise_std = 0.1\n",
    "n_test = 1000\n",
    "\n",
    "ns = np.array([400])\n",
    "ds = np.asarray(gammas*ns[0], dtype=int)\n",
    "\n",
    "mdl_practice = np.zeros((num_trials,  len(ds), len(ns)))\n",
    "mdl_theory = np.zeros((num_trials, len(ds), len(ns)))\n",
    "mdl_theory_mean = np.zeros((num_trials, len(ds), len(ns)))\n",
    "mdl_theory_lam  = np.zeros((num_trials, len(ds), len(ns)))\n",
    "mse = np.zeros((num_trials, len(ds), len(ns)))\n",
    "tnorm = np.zeros((num_trials, len(ds), len(ns)))\n",
    "eigsum = np.zeros((num_trials, len(ds), len(ns)))\n",
    "\n",
    "mse_mdl = np.zeros((num_trials, len(ds), len(ns))) ## this is mdl's\n",
    "\n",
    "mse_ols = np.zeros((num_trials, len(ds), len(ns)))\n",
    "mse_ridge = np.zeros((num_trials, len(ds), len(ns)))\n",
    "\n",
    "norm_beta = np.zeros((len(ds), len(ns)))\n",
    "epow = 0.\n",
    "for k, n in enumerate(ns):\n",
    "    ds = np.asarray(np.maximum(2, gammas*n), dtype=int)\n",
    "\n",
    "    print('n=%d'%n)\n",
    "    for j, d in enumerate(ds):\n",
    "        evals = np.array([1/i**epow for i in range(1, d+1)])\n",
    "        print('d=%d'%d)\n",
    "        X_train = npr.randn(n, d) / np.sqrt(n) @ np.diag(evals)\n",
    "        X_test = npr.randn(n_test, d) @ np.diag(evals)\n",
    "        eigenvals, eigenvecs = scl.eigh(X_train.T @ X_train)\n",
    "        \n",
    "        w = 1./np.arange(1, d+1)**0.25 / np.maximum(np.sqrt(1e-2), np.sqrt(np.maximum(0., eigenvals)))\n",
    "#         w /= npl.norm(w)\n",
    "\n",
    "        beta_star = eigenvecs.T @ w\n",
    "\n",
    "        norm_beta[j, k] = npl.norm(beta_star)\n",
    "        for i in range(num_trials):\n",
    "        \n",
    "            y_train = X_train @ beta_star + noise_std * npr.randn(n)\n",
    "            y_test =  X_test @ beta_star # + noise_std *  npr.randn(n_test)\n",
    "\n",
    "            var = noise_std**2\n",
    "            mdl_theory[i, j, k] = np.sum(np.log(1+ eigenvals * w**2 / var)) / n\n",
    "            \n",
    "            m = RidgeCV(fit_intercept=False, alphas=np.logspace(-5, 6, num=100, base=10))\n",
    "            m.fit(X_train, y_train)            \n",
    "            mse_ridge[i, j, k] = npl.norm(y_test - X_test @ m.coef_)**2 /  n_test\n",
    "                        \n",
    "            ols = LinearRegression(fit_intercept=False)\n",
    "            ols.fit(X_train, y_train)\n",
    "\n",
    "            mse_ols[i, j, k] = npl.norm(y_test - X_test @ ols.coef_)**2 / n_test\n",
    "            \n",
    "            "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Dump Results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {},
   "outputs": [],
   "source": [
    "file_name = 'remove_double_descent.txt'\n",
    "results = {'ns': ns, 'ds': ds, \n",
    "           'gammas': gammas,\n",
    "           'noise_var': noise_std**2,\n",
    "           'num_trials': num_trials, \n",
    "           'norm_beta': norm_beta, \n",
    "           'mse_ridge': mse_ridge, \n",
    "           'mse_ols':mse_ols, \n",
    "           'mdl_theory': mdl_theory,\n",
    "          }\n",
    "file = open(file_name, 'wb')\n",
    "pickle.dump(results, file)\n",
    "file.close()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Load Results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {},
   "outputs": [],
   "source": [
    "file_name = 'remove_double_descent.txt'\n",
    "file = open(file_name, 'rb')\n",
    "results = pickle.load(file)\n",
    "file.close()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "ns = results[\"ns\"]\n",
    "gammas = results[\"gammas\"]\n",
    "mdl_theory = results[\"mdl_theory\"]\n",
    "norm_beta = results[\"norm_beta\"]\n",
    "mse_ridge = results[\"mse_ridge\"]\n",
    "mse_ols = results[\"mse_ols\"]\n",
    "\n",
    "for k, n in enumerate(ns):\n",
    "    plt.figure(figsize=[8, 6])\n",
    "    ds = np.asarray(np.maximum(2, gammas*n), dtype=int)\n",
    "    plt.loglog(ds/n, mse_ols.mean(0)[:, k], color=cs[0], linestyle='--', linewidth=4., marker='d',  markersize=10, label='OLS Test-MSE')\n",
    "    plt.loglog(ds/n, mse_ridge.mean(0)[:, k], color=cs[2], linestyle=':', linewidth=4., marker='s',  markersize=8,  label='Ridge Test-MSE')\n",
    "    plt.loglog(ds/n, mdl_theory.mean(0)[:, k], color=cs[1], linestyle='-',linewidth=3.,  label='MDL-COMP')\n",
    "    plt.loglog(ds/n, norm_beta[:, k]**2, color=cs[3], linestyle=\"-.\", linewidth=3.,  label=r'$\\Vert \\theta_* \\Vert_2^2$')\n",
    "    plt.xlabel('d/n')\n",
    "    plt.legend(frameon=False, fontsize=20)\n",
    "    plt.grid(True, alpha=0.5, linewidth=1.)\n",
    "    plt.tight_layout()\n",
    "    plt.savefig('remove_double_descent_part1_n_%d.pdf'%n)\n",
    "    plt.show()\n",
    "\n",
    "for k, n in enumerate(ns):\n",
    "    plt.figure(figsize=[8, 6])\n",
    "    plt.semilogy(mdl_theory.mean(0)[:, k], mse_ols.mean(0)[:, k], color=cs[0], linestyle='--', linewidth=4., marker='d',  markersize=10, label='OLS Test-MSE')\n",
    "    plt.semilogy(mdl_theory.mean(0)[:, k], mse_ridge.mean(0)[:, k], color=cs[2], linestyle=':', linewidth=3., marker='s',  markersize=8,  label='Ridge Test-MSE')\n",
    "    plt.xlabel('MDL-COMP')\n",
    "    plt.xticks([0.5, 1., 1.5, 2.], [0.5, 1., 1.5, 2.])\n",
    "    plt.legend(frameon=False, fontsize=20, loc='center left')    \n",
    "    plt.tight_layout()\n",
    "    plt.grid(True, alpha=0.5, linewidth=1.)\n",
    "    plt.savefig('remove_double_descent_part2_n_%d.pdf'%n)\n",
    "    plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 167,
   "metadata": {},
   "outputs": [],
   "source": [
    "# thm3_bnds_new = dict()\n",
    "# for k, n in enumerate(ns):\n",
    "#     ds = np.asarray(n*gammas, dtype=int)\n",
    "#     thm3_bnds_new[n] = thm3(ds/n, 1./var*norm_beta[:, k]**2/ds)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Just MDL Expressions with simple decay"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 128,
   "metadata": {},
   "outputs": [],
   "source": [
    "gammas = np.logspace(-2., 2., 11)\n",
    "\n",
    "noise_std = 0.1\n",
    "\n",
    "ns = np.array([400])\n",
    "ds = np.asarray(gammas*ns[0], dtype=int)\n",
    "\n",
    "epows = np.linspace(0., .5, 2)\n",
    "thm_mdl_comp = np.zeros((len(ds), len(epows)))\n",
    "\n",
    "thm_mdl_comp_2 = np.zeros((len(ds), len(epows)))\n",
    "C = 100.\n",
    "\n",
    "for k, epow in enumerate(epows):\n",
    "    for j, d in enumerate(ds):\n",
    "        ew = 1./np.arange(1, d+1)**epow\n",
    "        ew /= np.sum(ew)\n",
    "        ew *= d\n",
    "        thm_mdl_comp[j, k] = np.sum(np.log(1+ ew[:min(d, ns[0])])) / ns[0]\n",
    "        \n",
    "        ew/= d\n",
    "        thm_mdl_comp_2[j, k] = np.sum(np.log(1+ 100*ew[:min(d, ns[0])])) / ns[0]\n",
    "        "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 129,
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib as mpl\n",
    "label_size = 20\n",
    "mpl.rcParams['xtick.labelsize'] = label_size \n",
    "mpl.rcParams['ytick.labelsize'] = label_size \n",
    "mpl.rcParams['axes.labelsize'] = label_size\n",
    "mpl.rcParams['axes.titlesize'] = 14\n",
    "mpl.rcParams['figure.titlesize'] = 10\n",
    "# mpl.rcParams['lines.markersize'] = 8\n",
    "mpl.rcParams['grid.linewidth'] = 2.5\n",
    "mpl.rcParams['legend.fontsize'] = 12\n",
    "# pylab.rcParams['xtick.major.pad']=5\n",
    "# pylab.rcParams['ytick.major.pad']=5"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 130,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "lss = ['-', '--', '-.', ':']\n",
    "mss = ['s', 'd', 'o', '^']\n",
    "plt.figure(figsize=[8, 6])\n",
    "for k, epow in enumerate(epows):\n",
    "    if epow==0:\n",
    "        plt.loglog(gammas, thm_mdl_comp[:, k], label=r'$(\\alpha, S_d)= (%d, d)$'%epow, marker=mss[k], linestyle=lss[k], linewidth=4., markersize=10)\n",
    "        plt.loglog(gammas, thm_mdl_comp_2[:, k], label=r'$(\\alpha, S_d)= (%d, 100)$'%epow, linestyle=lss[k+2],  marker=mss[k+2],  linewidth=4., markersize=10)\n",
    "    else:\n",
    "        plt.loglog(gammas, thm_mdl_comp[:, k], label=r'$(\\alpha, S_d)= (%.1f, d)$'%epow, marker=mss[k], linestyle=lss[k], linewidth=4., markersize=10)\n",
    "        plt.loglog(gammas, thm_mdl_comp_2[:, k], label=r'$(\\alpha, S_d)= (%.1f, 100)$'%epow, linestyle=lss[k+2],  marker=mss[k+2],  linewidth=4., markersize=10)\n",
    "# plt.axvline(1., color='k', alpha=0.5, linewidth=2., linestyle=':')\n",
    "plt.ylabel('MDL-COMP')\n",
    "plt.grid(True, alpha=0.5, linewidth=1.)\n",
    "plt.ylim([1e-3, 3])\n",
    "plt.legend(loc='lower center', frameon=False, fontsize=20)\n",
    "plt.xlabel('d/n')\n",
    "plt.tight_layout()\n",
    "plt.savefig('theory_decay_mdl.pdf')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## MDL Expressions with Random Matrix with Decay"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "gammas_new = np.logspace(-1., 1., 5)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "gammas_new = np.logspace(-1., 1., 5)\n",
    "\n",
    "noise_std = 0.1\n",
    "\n",
    "ns = np.array([100])\n",
    "n = ns[0]\n",
    "ds = np.asarray(gammas_new*ns[0], dtype=int)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "  0%|          | 0/5 [00:00<?, ?it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "d=10\n",
      "d=31\n",
      "d=100\n",
      "d=316\n",
      "d=1000\n",
      "d=10\n",
      "d=31\n",
      "d=100\n",
      "d=316\n",
      "d=1000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 20%|██        | 1/5 [00:00<00:02,  1.84it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "d=10\n",
      "d=31\n",
      "d=100\n",
      "d=316\n",
      "d=1000\n",
      "d=10\n",
      "d=31\n",
      "d=100\n",
      "d=316\n",
      "d=1000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 40%|████      | 2/5 [00:01<00:01,  1.85it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "d=10\n",
      "d=31\n",
      "d=100\n",
      "d=316\n",
      "d=1000\n",
      "d=10\n",
      "d=31\n",
      "d=100\n",
      "d=316\n",
      "d=1000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 60%|██████    | 3/5 [00:01<00:01,  1.85it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "d=10\n",
      "d=31\n",
      "d=100\n",
      "d=316\n",
      "d=1000\n",
      "d=10\n",
      "d=31\n",
      "d=100\n",
      "d=316\n",
      "d=1000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 80%|████████  | 4/5 [00:02<00:00,  1.86it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "d=10\n",
      "d=31\n",
      "d=100\n",
      "d=316\n",
      "d=1000\n",
      "d=10\n",
      "d=31\n",
      "d=100\n",
      "d=316\n",
      "d=1000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 5/5 [00:02<00:00,  1.87it/s]\n"
     ]
    }
   ],
   "source": [
    "npr.seed(0)\n",
    "num_trials = 5\n",
    "rmt_mdl_comp = np.zeros((num_trials,len(ds), len(epows)))\n",
    "\n",
    "rmt_mdl_comp_2 = np.zeros((num_trials, len(ds), len(epows)))\n",
    "\n",
    "for p in tqdm(range(num_trials)):\n",
    "    for k, epow in enumerate(epows):\n",
    "        for j, d in enumerate(ds):\n",
    "            evals = np.array([1/i**epow for i in range(1, d+1)])\n",
    "            print('d=%d'%d)\n",
    "            X_train = npr.randn(n, d) / np.sqrt(n) @ np.diag(evals)\n",
    "            eigenvals, eigenvecs = scl.eigh(X_train.T @ X_train)\n",
    "\n",
    "            beta_star = npr.randn(d)\n",
    "            beta_star /= npl.norm(beta_star)\n",
    "\n",
    "            w = eigenvecs.T @ beta_star\n",
    "\n",
    "            rmt_mdl_comp[p, j, k] = np.sum(np.log(1+100* eigenvals * w**2/noise_std**2)) / ns[0]\n",
    "            rmt_mdl_comp_2[p, j, k] = np.sum(np.log(1+ d*eigenvals * w**2/noise_std**2)) / ns[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 132,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "lss = ['-', '--', '-.', ':']\n",
    "mss = ['s', 'd', 'o', '^']\n",
    "\n",
    "plt.figure(figsize=[8, 6])\n",
    "for k, epow in enumerate(epows):\n",
    "    if epow==0:\n",
    "        plt.loglog(gammas_new, rmt_mdl_comp_2.mean(0)[:, k], label=r'$(\\alpha, R_d)= (%d, d)$'%epow, marker=mss[k], linestyle=lss[k], linewidth=4., markersize=10)\n",
    "        plt.loglog(gammas_new, rmt_mdl_comp.mean(0)[:, k], label=r'$(\\alpha, R_d)= (%d, 100)$'%epow, \n",
    "                   linestyle=lss[k+2],  marker=mss[k+2],  linewidth=4., markersize=10)\n",
    "    else:\n",
    "        plt.loglog(gammas_new, rmt_mdl_comp_2.mean(0)[:, k], label=r'$(\\alpha, R_d)= (%.1f, d)$'%epow, marker=mss[k], \n",
    "                   linestyle=lss[k], linewidth=4., markersize=10)\n",
    "        plt.loglog(gammas_new, rmt_mdl_comp.mean(0)[:, k], label=r'$(\\alpha, R_d)= (%.1f, 100)$'%epow, linestyle=lss[k+2],  \n",
    "                   marker=mss[k+2],  linewidth=4., markersize=10)\n",
    "# plt.axvline(1., color='k', alpha=0.5, linewidth=2., linestyle=':')\n",
    "plt.grid(True, alpha=0.5, linewidth=1.)\n",
    "plt.ylabel('MDL-COMP')\n",
    "plt.ylim([5e-2, 1e1])\n",
    "plt.legend(frameon=False, fontsize=20, loc='lower center')\n",
    "plt.xlabel('d/n')\n",
    "plt.tight_layout()\n",
    "plt.savefig('RMT_decay_mdl.pdf')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 116,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0.43242953, 0.74894956, 0.77634125, 0.5994712 , 0.36975855])"
      ]
     },
     "execution_count": 116,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rmt_mdl_comp.mean(0)[:, 1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 131,
   "metadata": {},
   "outputs": [],
   "source": [
    "file_name = 'mdl_complexities.txt'\n",
    "results = {\n",
    "            'gammas_new': gammas_new,\n",
    "           'rmt_mdl_comp': rmt_mdl_comp, \n",
    "           'rmt_mdl_comp_2': rmt_mdl_comp_2, \n",
    "           'thm_mdl_comp': thm_mdl_comp, \n",
    "           'thm_mdl_comp_2':thm_mdl_comp_2, \n",
    "           'epows': epows,\n",
    "            'r_ds': 'd, 100',\n",
    "            's_ds': 'd, 100'\n",
    "          }\n",
    "file = open(file_name, 'wb')\n",
    "pickle.dump(results, file)\n",
    "file.close()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## MDL New"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Once again"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 123,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "n=100\n",
      "d=10\n",
      "d=31\n",
      "d=100\n",
      "d=316\n",
      "d=1000\n",
      "d=3162\n"
     ]
    }
   ],
   "source": [
    "npr.seed(0)\n",
    "num_trials = 5\n",
    "\n",
    "gammas = np.hstack((np.logspace(-1, -0.1, 5),\n",
    "        np.logspace(-0.05, 0.1, 4),\n",
    "        np.logspace(0.2, 1., 5)))\n",
    "\n",
    "gammas = np.logspace(-1., 1.5, 6)\n",
    "# gammas = np.array([0.1, 0.5, 1., 2., 5.])\n",
    "noise_std = 1\n",
    "n_test = 1000\n",
    "\n",
    "ns = np.array([100])\n",
    "ds = np.asarray(gammas*ns[0], dtype=int)\n",
    "\n",
    "mdl_practice = np.zeros((num_trials,  len(ds), len(ns)))\n",
    "mdl_theory = np.zeros((num_trials, len(ds), len(ns)))\n",
    "mdl_theory_mean = np.zeros((num_trials, len(ds), len(ns)))\n",
    "mdl_theory_lam  = np.zeros((num_trials, len(ds), len(ns)))\n",
    "mse = np.zeros((num_trials, len(ds), len(ns)))\n",
    "tnorm = np.zeros((num_trials, len(ds), len(ns)))\n",
    "eigsum = np.zeros((num_trials, len(ds), len(ns)))\n",
    "\n",
    "mse_mdl = np.zeros((num_trials, len(ds), len(ns))) ## this is mdl's\n",
    "\n",
    "mse_ols = np.zeros((num_trials, len(ds), len(ns)))\n",
    "mse_ridge = np.zeros((num_trials, len(ds), len(ns)))\n",
    "\n",
    "norm_beta = np.zeros((len(ds), len(ns)))\n",
    "epow = 0.5\n",
    "for k, n in enumerate(ns):\n",
    "    ds = np.asarray(np.maximum(2, gammas*n), dtype=int)\n",
    "\n",
    "    print('n=%d'%n)\n",
    "    for j, d in enumerate(ds):\n",
    "        evals = np.array([1/i**epow for i in range(1, d+1)])\n",
    "        print('d=%d'%d)\n",
    "        X_train = npr.randn(n, d) / np.sqrt(n) @ np.diag(evals)\n",
    "        X_test = npr.randn(n_test, d) @ np.diag(evals)\n",
    "        eigenvals, eigenvecs = scl.eigh(X_train.T @ X_train)\n",
    "        \n",
    "        beta_star = npr.randn(d)\n",
    "        beta_star /= npl.norm(beta_star)\n",
    "        beta_star *= 10\n",
    "        \n",
    "        w = eigenvecs.T @ beta_star\n",
    "\n",
    "        norm_beta[j, k] = npl.norm(beta_star)\n",
    "        for i in range(num_trials):\n",
    "        \n",
    "            y_train = X_train @ beta_star + noise_std * npr.randn(n)\n",
    "            y_test =  X_test @ beta_star # + noise_std *  npr.randn(n_test)\n",
    "\n",
    "            var = noise_std**2\n",
    "            mdl_theory[i, j, k] = np.sum(np.log(1+ eigenvals * w**2 / var)) / n\n",
    "            \n",
    "            m = RidgeCV(fit_intercept=False, alphas=np.logspace(-5, 6, num=100, base=10))\n",
    "            m.fit(X_train, y_train)            \n",
    "            mse_ridge[i, j, k] = npl.norm(y_test - X_test @ m.coef_)**2 /  n_test\n",
    "                        \n",
    "            ols = LinearRegression(fit_intercept=False)\n",
    "            ols.fit(X_train, y_train)\n",
    "\n",
    "            mse_ols[i, j, k] = npl.norm(y_test - X_test @ ols.coef_)**2 / n_test\n",
    "            \n",
    "            "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 126,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "for k, n in enumerate(ns):\n",
    "    plt.figure(figsize=[8, 6])\n",
    "    ds = np.asarray(np.maximum(2, gammas*n), dtype=int)\n",
    "    plt.loglog(ds/n, mse_ols.mean(0)[:, k], color=cs[0], linestyle='--', linewidth=4., marker='d',  markersize=10, label='OLS Test-MSE')\n",
    "    plt.loglog(ds/n, mse_ridge.mean(0)[:, k], color=cs[2], linestyle=':', linewidth=4., marker='s',  markersize=8,  label='Ridge Test-MSE')\n",
    "    plt.loglog(ds/n, mdl_theory.mean(0)[:, k], color=cs[1], linestyle='-',linewidth=3.,  label='MDL-COMP')\n",
    "    plt.loglog(ds/n, norm_beta[:, k]**2, color=cs[3], linestyle=\"-.\", linewidth=3.,  label=r'$\\Vert \\theta_* \\Vert_2^2$')\n",
    "    plt.xlabel('d/n')\n",
    "    plt.legend(frameon=False, fontsize=20)\n",
    "    plt.grid(True, alpha=0.5, linewidth=1.)\n",
    "    plt.tight_layout()\n",
    "#     plt.savefig('remove_double_descent_part1_n_%d.pdf'%n)\n",
    "    plt.show()\n",
    "\n",
    "# for k, n in enumerate(ns):\n",
    "#     plt.figure(figsize=[8, 6])\n",
    "#     rmt_mdl_comp.mean(0)[:, 1]\n",
    "#     plt.semilogy(rmt_mdl_comp.mean(0)[:, 1], mse_ols.mean(0)[:, k], color=cs[0], linestyle='--', linewidth=4., marker='d',  markersize=10, label='OLS Test-MSE')\n",
    "#     plt.semilogy(rmt_mdl_comp.mean(0)[:, 1], mse_ridge.mean(0)[:, k], color=cs[2], linestyle=':', linewidth=3., marker='s',  markersize=8,  label='Ridge Test-MSE')\n",
    "\n",
    "# #     plt.semilogy(mdl_theory.mean(0)[:, k], mse_ols.mean(0)[:, k], color=cs[0], linestyle='--', linewidth=4., marker='d',  markersize=10, label='OLS Test-MSE')\n",
    "    \n",
    "# #     plt.semilogy(mdl_theory.mean(0)[:, k], mse_ridge.mean(0)[:, k], color=cs[2], linestyle=':', linewidth=3., marker='s',  markersize=8,  label='Ridge Test-MSE')\n",
    "#     plt.xlabel('MDL-COMP')\n",
    "# #     plt.xticks([0.5, 1., 1.5, 2.], [0.5, 1., 1.5, 2.])\n",
    "#     plt.legend(frameon=False, fontsize=20, loc='center left')    \n",
    "#     plt.tight_layout()\n",
    "#     plt.grid(True, alpha=0.5, linewidth=1.)\n",
    "# #     plt.savefig('remove_double_descent_part2_n_%d.pdf'%n)\n",
    "#     plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Fixed X"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "npr.seed(0)\n",
    "num_trials = 1\n",
    "gammas_new = np.logspace(-1., 1.5, 6)\n",
    "\n",
    "noise_std = .1\n",
    "\n",
    "ds = np.array([1000])\n",
    "d = ds[0]\n",
    "ns = np.asarray(d/gammas_new, dtype=int)\n",
    "\n",
    "theta_type = ['inc', 'dec', 'rand']\n",
    "\n",
    "norm_type = ['const', 'inc']\n",
    "\n",
    "fixed_mdl_comp = {} \n",
    "\n",
    "X_train = npr.randn(max(ns), d) @ np.diag(np.arange(1, d+1)**-0.5)\n",
    "\n",
    "\n",
    "for k, tt in enumerate(theta_type):\n",
    "    fixed_mdl_comp[tt] = {}\n",
    "\n",
    "    for i, nt in enumerate(norm_type):\n",
    "        fixed_mdl_comp[tt][nt] = np.zeros(len(ns))\n",
    "        \n",
    "        for j, n in enumerate(ns):\n",
    "            eigenvals, eigenvecs = scl.eigh(X_train[:n, :].T @ X_train[:n, :])\n",
    "            if tt == \"inc\":\n",
    "                tstar = 1.*np.arange(1, d+1)**0.5\n",
    "            elif tt == \"dec\":\n",
    "                tstar = (1.*np.arange(1, d+1)**0.5)[::-1]\n",
    "            elif tt == \"rand\":\n",
    "                tstar = npr.randn(d)\n",
    "            \n",
    "            tstar /= npl.norm(tstar)\n",
    "            if nt == \"const\":\n",
    "                tstar *= 1.\n",
    "            elif nt == \"inc\":\n",
    "                tstar *= np.sqrt(d)\n",
    "            elif nt == \"dec\":\n",
    "                tstar /= np.sqrt(d)\n",
    "        \n",
    "            w = eigenvecs.T @ tstar\n",
    "\n",
    "            fixed_mdl_comp[tt][nt][j]  = 1.*np.sum(np.log(1+eigenvals * w**2/noise_std**2)) / n\n",
    "            \n",
    "        plt.loglog(gammas_new, fixed_mdl_comp[tt][nt], label=tt+' '+nt)\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n",
    "# Dump Results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "file_name = 'SOMETHING_NEW.txt'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "results = {'ns': ns, 'ds': ds, \n",
    "           'num_trials': num_trials, \n",
    "           'beta_norm': npl.norm(beta_star), \n",
    "           'noise_var': noise_std**2,\n",
    "           'gammas': gammas,\n",
    "           'mse': mse, \n",
    "           'eigsum': eigsum, \n",
    "           'tnorm':tnorm,\n",
    "           'mdl_practice': mdl_practice, \n",
    "\n",
    "           'mdl_theory':mdl_theory, \n",
    "           'mdl_thm_3': thm3_bnds,\n",
    "           \n",
    "#            'mdl_theory_lam': mdl_theory_lam, \n",
    "#            'mdl_theory_mean': mdl_theory_mean,\n",
    "\n",
    "          }\n",
    "file = open(file_name, 'wb')\n",
    "pickle.dump(results, file)\n",
    "file.close()"
   ]
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}