bdt.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Welcome to JupyROOT 6.22/02\n"
     ]
    }
   ],
   "source": [
    "import numpy as np\n",
    "import scipy\n",
    "import pandas as pd\n",
    "import seaborn as sn\n",
    "from matplotlib import pyplot as plt\n",
    "\n",
    "from sklearn.tree import DecisionTreeRegressor\n",
    "\n",
    "import xgboost as xgb\n",
    "from sklearn.metrics import mean_squared_error\n",
    "\n",
    "import sklearn.tree\n",
    "\n",
    "from IPython.display import Image\n",
    "\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn import preprocessing\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "\n",
    "from sklearn.model_selection import KFold\n",
    "\n",
    "from sklearn.metrics import r2_score\n",
    "\n",
    "import uproot\n",
    "import ROOT\n",
    "from array import array"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "#from numpy.random import seed\n",
    "#import random as python_random\n",
    "#number = 1\n",
    "#seed(number)\n",
    "#python_random.seed(number)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "#variables_test = ([\"n30\", \"n35\", \"n40\", \"n45\", \"n50\", \"n55\", \"n60\", \"n65\", \"n70\", \"n75\", \"n80\", \"nav\", \"c30\", \n",
    "#\"c35\", \"c40\", \"c45\", \"c50\", \"c55\", \"c60\", \"c65\", \"c70\", \"c75\", \"c80\", ])\n",
    "#branches = [ \"energy\", \"eecal\", \"ehcal\", \"nhits\", \"nhits_ecal\", \"nhits_hcal\", \"time\", \"etime\", \"htime\", \"radius\", \n",
    "#\"eradius\", \"hradius\", \"elayer\", \"hlayer\", \"energy_hit\", \"energy_hit_ecal\", \"energy_hit_hcal\", \n",
    "#\"energy_hit_ecal_max\", \"energy_hit_hcal_max\", \"muon_energy\", \"muon_energy_hit\", \"muon_nhits\", \"muon_time\", \n",
    "#\"muon_layer\" ]\n",
    "\n",
    "#branches = [ \"energy\", \"eecal\", \"ehcal\", \"nhits\", \"nhits_ecal\", \"nhits_hcal\", \"radius\", \"eradius\", \"hradius\", \n",
    "#\"time\", \"etime\", \"htime\", \"elayer\", \"hlayer\", \"energy_hit\", \"energy_hit_ecal\", \"energy_hit_hcal\", \n",
    "#\"energy_hit_ecal_max\", \"energy_hit_hcal_max\", \"muon_energy\", \"muon_energy_hit\", \"muon_nhits\", \"muon_time\", \n",
    "#\"muon_layer\" ]\n",
    "\n",
    "#branches = [\"energy\", \"ehcal\", \"eecal\", \"nhits\", \"nhits_ecal\", \"nhits_hcal\", \"htime\", \"radius\", \"hradius\", \n",
    "#            \"hlayer\", \"c30\", \"energy_hit\", \"energy_hit_hcal\", \"muon_energy\", \"muon_energy_hit\", \"muon_nhits\", \n",
    "#            \"muon_time\", \"muon_layer\"]\n",
    "\n",
    "branches = [\"eecal\", \"ehcal\", \"energy_hit_ecal\", \"energy_hit_hcal\", \"muon_energy\", \"muon_energy_hit\", \n",
    "           \"nhits_ecal\", \"nhits_hcal\", \"etime\", \"htime\", \"eradius\", \"hradius\", \"elayer\", \"hlayer\",\n",
    "           \"muon_nhits\", \"muon_time\", \"muon_layer\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# ./gun_k0L_allgev_FTFP_BERT_90000evt_ILD_l5_v02steel.root\n",
    "\n",
    "fTreeName = \"tree\"\n",
    "\n",
    "data = uproot.pandas.iterate(\"./gun_k0L_allgev_FTFP_BERT_260000evt_ILD_l5_v02steel.root\", fTreeName, branches)\n",
    "data = pd.concat(list(data))\n",
    "\n",
    "data_mc = uproot.pandas.iterate(\"./gun_k0L_allgev_FTFP_BERT_260000evt_ILD_l5_v02steel.root\", fTreeName, [\"mc_energy\"])\n",
    "data_mc = pd.concat(list(data_mc))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "x_train, x_test, y_train, y_test = train_test_split(data, data_mc, train_size = 0.9)\n",
    "#x_train, x_test, y_train, y_test = train_test_split(data_scaled, data_mc_scaled, train_size = 0.9)\n",
    "#x = data[data.columns[:2]]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "#def loss3(y_true, y_pred):\n",
    "#    residual = (y_true - y_pred).astype(\"float\")\n",
    "#    grad = np.where(residual<0, -2*10.0*residual, -2*residual)\n",
    "#    hess = np.where(residual<0, 2*10.0, 2.0)\n",
    "#    return grad, hess\n",
    "\n",
    "\n",
    "\n",
    "#def loss3(y_true, y_pred):\n",
    "#    return k.mean( k.square( (y_pred - y_true) ) / ( a * y_true) )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "#xg_reg = xgb.XGBRegressor(objective = loss3, colsample_bytree = 0.3, learning_rate = 0.1,\n",
    "#                max_depth = 30, alpha = 10, n_estimators = 100)\n",
    "\n",
    "xg_reg = xgb.XGBRegressor(colsample_bytree = 0.3, learning_rate = 0.1,\n",
    "                max_depth = 20, alpha = 10, n_estimators = 10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "history = xg_reg.fit(x_train,y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "preds = xg_reg.predict(x_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RMSE: 16.790770\n"
     ]
    }
   ],
   "source": [
    "rmse = np.sqrt(mean_squared_error(y_test, preds))\n",
    "print(\"RMSE: %f\" % (rmse))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_preds = pd.Series(preds, name='y_preds', dtype='float32')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:ylabel='Frequency'>"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "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": [
    "y_preds.plot(kind='hist')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAckAAAEWCAYAAAANV2yLAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuMCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy86wFpkAAAACXBIWXMAAAsTAAALEwEAmpwYAABVAUlEQVR4nO3dd3xUVfrH8c9DkSogUqRIRzqMgAiKGkWKgCJ2RBEREbvrqqCiIrsqi+KKghULIoJI11WEHxBFFKWFIkjZJSsEliY1BAjh+f1xT4ZJmEkmkGSSyfN+veaVe89t52Q0h9u+R1QVY4wxxpyqUKQrYIwxxuRV1kkaY4wxIVgnaYwxxoRgnaQxxhgTgnWSxhhjTAjWSRpjjDEhWCdpjDkjIvKMiIyNdD2MyQli70kaEzkiEg9UBlICii9Q1W1nuM/+qvp/Z1a7/EdEhgL1VPWOSNfFRAc7kzQm8q5V1dIBn9PuILODiBSJ5PFPV36tt8nbrJM0Jg8SkbIi8qGIbBeRBBH5u4gUdsvqish8EdkjIrtFZIKIlHPLxgM1gK9E5JCIPCUiMSKyNd3+40Xkajc9VESmiMhnInIA6JvR8YPUdaiIfOama4mIisjdIrJFRPaKyEARuUhEVonIPhEZHbBtXxFZJCKjRWS/iPwuIh0CllcVkVki8qeIbBKRe9MdN7DeA4FngFtd21e69e4WkXUiclBE/iMi9wXsI0ZEtorIX0Vkp2vv3QHLS4jISBH5r6vfjyJSwi1rKyI/uTatFJGY0/iqTR5nnaQxedMnwHGgHnAh0Ano75YJ8ApQFWgEnA8MBVDVO4E/OHl2OiLM4/UApgDlgAmZHD8cFwP1gVuBN4BngauBJsAtInJFunX/DVQAXgCmiUh5t2wSsNW19SbgZRG5KkS9PwReBr5wbW/h1tkJdAfKAHcD/xSRlgH7OA8oC1QD7gHGiMg5btlrQCvgEqA88BRwQkSqAf8C/u7KnwCmikjFLPyOTD5gnaQxkTfDnY3sE5EZIlIZ6Ao8pqqJqroT+CdwG4CqblLVuap6VFV3Aa8DV4TefVh+VtUZqnoCrzMJefww/U1Vj6jqHCARmKiqO1U1AViI1/Gm2gm8oarJqvoFsB7oJiLnA5cCg9y+4oCxQJ9g9VbVpGAVUdV/qeq/1fM9MAe4LGCVZGCYO/43wCGggYgUAvoBj6pqgqqmqOpPqnoUuAP4RlW/cceeCyx1vzcTRewavjGRd33gQzYi0gYoCmwXkdTiQsAWt7wyMArvD/3ZbtneM6zDloDpmhkdP0w7AqaTgsyXDphP0LRPEP4X78yxKvCnqh5Mt6x1iHoHJSLX4J2hXoDXjpLA6oBV9qjq8YD5w65+FYDieGe56dUEbhaRawPKigILMquPyV+skzQm79kCHAUqpPvjneplQIFmqvqniFwPjA5Ynv6R9US8jgEAd28x/WXBwG0yO352qyYiEtBR1gBmAduA8iJydkBHWQNICNg2fVvTzItIMWAq3tnnTFVNFpEZeJesM7MbOALUBVamW7YFGK+q956ylYkqdrnVmDxGVbfjXRIcKSJlRKSQe1gn9ZLq2XiXBPe7e2NPptvFDqBOwPwGoLiIdBORosAQoNgZHD+7VQIeEZGiInIz3n3Wb1R1C/AT8IqIFBeR5nj3DD/LYF87gFruUinAWXht3QUcd2eVncKplLv0/BHwunuAqLCItHMd72fAtSLS2ZUXdw8BVc96801eZp2kMXlTH7w/8GvxLqVOAaq4ZS8CLYH9eA+PTEu37SvAEHeP8wlV3Q88gHc/LwHvzHIrGcvo+NntF7yHfHYDLwE3qeoet6wXUAvvrHI68EIm739+6X7uEZHl7gz0EWAyXjtuxztLDdcTeJdmlwB/Av8ACrkOvAfe07S78M4sn8T+pkYdCxMwxkSMiPTFCz5oH+m6GBOM/avHGGOMCcE6SWOMMSYEu9xqjDHGhGBnksYYY0wI9p5kFClXrpzWq1cv0tXIcYmJiZQqVSrS1cgV1tboVFDamh/auWzZst2qGjJO0DrJKFK5cmWWLl0a6WrkuNjYWGJiYiJdjVxhbY1OBaWt+aGdIvLfjJbb5VZjjDEmBOskjTHGmBCskzTGGGNCsE7SGGOMCcE6SWOMMSYE6ySNMcaYEKyTNMYYkyf069ePSpUq0bRpU3/Zn3/+SceOHalfvz4dO3Zk715vfPGZM2fSvHlzfD4frVu35scff/RvU7hwYXw+Hz6fj+uuu85fftlll/nLq1atyvXXX59pnayTPA0iUktE1gQp/0REbopEnYwxJr/r27cvs2fPTlM2fPhwOnTowMaNG+nQoQPDhw8HoEOHDqxcuZK4uDg++ugj+vfv79+mRIkSxMXFERcXx6xZJ0dGW7hwob+8Xbt23HDDDZnWyTrJfERELPzBGBO1Lr/8csqXL5+mbObMmdx1110A3HXXXcyYMQOA0qVLIyKAl+yTOh2OAwcOMH/+/LDOJO2P7ukrLCIfAJfgDWTbI3ChiDwPXAuUwBtd/T680eK/VNWWbp36wBeq2lJEWgGvA6XxBp/tq6rbRSQWiAPaAxOBkaEqlJScQq3B/8rONuZJf212nL4FoJ1gbY1WBaWtWWln/PBuQct37NhBlSreeN/nnXceO3bs8C+bPn06Tz/9NDt37uRf/zp5nCNHjtC6dWuKFCnC4MGDT+kMZ8yYQYcOHShTpkym9bJO8vTVB3qp6r0iMhm4Md3y0ao6DEBExgPdVfUrEdkvIj5VjQPuBj4WkaLAW0APVd0lIrfijdDez+3rLFVtHawSIjIAGABQoUJFnm92PJubmfdULuH9z1cQWFujU0Fpa1baGRsbC8D//vc/EhMT/fPHjx/3TwOkpKT458855xzeffddVq5cyUMPPcTIkd45xMSJE6lYsSLbtm1j4MCBJCYmUq1aNf8+xowZQ9euXdPsNxTrJE/fZtfRASwDaqVbfqWIPAWUBMoDvwFfAWOBu0XkceBWoA3QAGgKzHWXDAoD2wP29UWoSqjq+8D7ADXq1NORq6P/K/1rs+MUhHaCtTVaFZS2ZqWd8b1jvJ/x8ZQqVcqf+VqtWjUaNGhAlSpV2L59O1WrVj0lDzYmJoZRo0bRtGlTKlSokGbZnDlzKFasmH+b3bt3s2nTJgYNGkTx4sUzrVf0f0s552jAdAreZVUARKQ48DbQWlW3iMhQIPXbmAq8AMwHlqnqHhGpCvymqu1CHCsxnAqVKFqY9SEuWUST2NhY//9Q0c7aGp0KSluzo53XXXcd48aNY/DgwYwbN44ePbw7W5s2baJu3bqICMuXL+fo0aOce+657N27l5IlS1KsWDF2797NokWLeOqpp/z7mzJlCt27dw+rgwTrJHNK6m9/t4iUBm4CpgCo6hER+Q54B7jHrbceqCgi7VT1Z3f59QJV/S23K26MMZHSq1cvYmNj2b17N9WrV+fFF19k8ODB3HLLLXz44YfUrFmTyZMnAzB16lQ+/fRTihYtSokSJfjiiy8QEdatW8d9991HoUKFOHHiBIMHD6Zx48b+Y0yaNInBgweHXSfrJHOAqu5zD/WsAf4HLEm3ygSgJzDHrX/MvTrypoiUxfte3sC7RGuMMQXCxIkTg5bPmzfvlLJBgwYxaNCgU8ovueQSVq9eHfIY4dyHDGSd5GlQ1Xi8e4ip868FWWcIMCTELtoDH6tqSsD6ccDlQfYTc2a1NcYYc7qsk8xlIjIdqAtcFem6GGOMyZh1krlMVXtGug7GGGPCY4k7xhiTDdavX+/PBfX5fJQpU4Y33ngDgLfeeouGDRvSpEkT/5OWc+fOpVWrVjRr1oxWrVoxf/58/75iYmJo0KCBf187d+6MRJMMdiZpjDHZokGDBsTFxQHeC+/VqlWjZ8+eLFiwgJkzZ7Jy5UqKFSvGzp07Wbt2LRUqVOCrr76iatWqrFmzhs6dO5OQkODf34QJE2jdOmiGiMlFdiaZjojEiMjXIZaNFZHGbvqZSNfHGJM3zZs3j7p161KzZk3eeecdBg8eTLFixQCoVKkSABdeeCFVq1YFoEmTJiQlJXH06NGQ+zSRETVnkiJSRFVzNOdJVfsHzD4DvJyTx8sqy26NPtbW/CMwe3TSpEn06tULgA0bNrBw4UKeffZZihcvzmuvnfIwPFOnTqVly5b+jhTg7rvvpnDhwtx4440MGTIkSwHeJvtE7ExSRO4QkV9FJE5E3hORwiJySEReEpGVIrJYRCq7dSuKyFQRWeI+l7ryoSIyXkQWAePdenNF5Dd31vdfEakgIsNE5LGAY78kIo9mUL3SIjJFRH4XkQni/usUkVgRaS0iw4ESru4TstJGV95FRJa7ds5zZW1E5GcRWSEiP4lIgzP9HRtjct+xY8eYNWsWN998M+Blj/75558sXryYV199lVtuuQVV9a//22+/MWjQIN577z1/2YQJE1i9ejULFy5k4cKFjB8/PtfbYTwROZMUkUZ4uaWXqmqyiLwN9AZKAYtV9VkRGQHcC/wdGAX8U1V/FJEawHdAI7e7xkB7VU0SkdHAfFV9RUS6cDLR5iNgGvCGiBQCbsPLTA3lQqAJsA1YBFwK+Ef0VNXBIvKQqvqy2kYR+Rb4ALhcVTeLSOq4ML8Dl6nqcRG5Gu8sNX1oerDjWMB5FLO25h+pL6n/+OOP1K5dm3Xr1rFu3TpKlixJnTp1+P777wGvE922bRuxsbHs2rWLxx9/nKeeeootW7awZcsW//42btwIQMuWLZk+fTo1atTI9TadqUOHDmX55f28JlKXWzsArYAl7iStBLATOAak3n9bBnR001cDjQMuN5RxcW8As1Q1yU23x0uyQVVni8heNx0vIntE5EKgMrBCVfdkUL9fVXUrgIjE4YWX/5jB+llpY1vgB1Xd7Or2p1u/LDDODZ+lQNFwDhIYcN6gQQN9uHePTLbI/2JjY7klXcBxtLK25j/vvvsuDzzwgD9Qu1+/fmzbto2YmBg2bNhAoUKFqFq1Kj6fjyuuuIJRo0alGfz3+PHj7Nu3jwoVKpCcnMzo0aPp3LnzKaHe+UFsbGy+rHegSHWSAoxT1afTFIo8oSevQ6Rwsn6FgLaqeiTd+hBm+Dfe6Bt9gfPwziwzkj68/HR+T6HaeG2I9f8GLFDVniJSC4g9jWMaYyIoMTGRuXPnprl02q9fP/r160fTpk0566yzGDduHCLC6NGj2bRpE8OGDWPYsGGAN2JFqVKl6Ny5M8nJyaSkpHD11Vdz7733RqpJBV6kOsl5wEwR+aeq7nSXHM/OYP05wMPAqwAB4zGmtwi4BfiHiHQCzglYNh0YhneGdvuZN4FkESmqqskhlodq42LgbRGpnXq51Z1NlsUbvBm8ztwYk8+UKlWKPXvSXqQ666yz+Oyzz9KUxcbGMmTIEIYMCZ5cuWzZshyro8maiDy4o6pr8XJN54jIKmAuUCWDTR4BWovIKhFZCwwMsd6LQCcRWQPcjBcuftAd8xiwAJgcmJl6Bt4HVoV6cCdUG1V1F949xGkispKTY0WOAF4RkRVE0VPHxhiTn0Xsj7GqfsGpgwmXDlg+hZPDS+3Gewgm/T6GpivaD3R2D7+0Ay5S1aMA7oGdtnidZ0b1iiXgUqeqPhQwHRMwPQg4NYI+7b6CtRFV/Rb4Nl3Zz8AFAUVDgtXHGGNM7om2M5YawGTXIR7DezoWFwDwNTBdVTdGsH7GGGPykahK3FHVjap6oaq2UNWLVHWJK1+rqnVU9a+p64pIM/f+YuDnl6weU0TODbKfOBE5NzvbZozJOzLKaQUYOXIkIsLu3bsBmDlzJs2bN8fn89G6des04x0+9dRTNGnShEaNGvHII4+keYfSRF60nUmGTVVXA75s2M+e9PsRkWcCXzERkZ9U9ZIzPZYxJm8IldMKsGXLFubMmZPmvcYOHTpw3XXXISKsWrWK7t278/DDD/PTTz+xaNEiVq1aBUD79u35/vvv8/1rE9Ekqs4k85A0ua7WQRoTvQJzWgH+8pe/MGLEiDQxcqVLl/bPJyYm+qdFhCNHjnDs2DGOHj1KcnIylStXzv1GmJAK7JlkdhGRO/Cevj0L+AU4gIusA35T1d4ickhVS4tIDN4TuPuAZsBkYDXwKF7YwPWq+m8RqQi8i3ePFeAxVV2UWV0suzX6WFvznsCMVkib0zpz5kyqVatGixYtTtlu+vTpPP300+zcuZO//e1vALRr144rr7ySKlWqoKo89NBDNGrU6JRtTeTYmeQZSBc958MLHlgNJKmqT1V7B9msBd4rLI2AO4ELVLUNXtjBw26d1Bi+i/Ci6cbmaEOMMaclMKf18OHDvPzyy/5ggPR69uzJ77//zowZM/joIy/PZNOmTaxbt46tW7eSkJDA/PnzWbhwYW42wWTCziTPTKjouYwsUdXtACLyb7ygBPA61yvddNAYPlU9lH5nlt0a3ayteU9gFmlgTut//vMfNmzYQIMG3tgEu3btokmTJrzzzjuUL18+zT4SEhKYOXMm3377LZUrV2bp0qUANGzYkM8++4yUlOx4lTvyoiG7FVW1z2l+8M78XglSfijYPBADfB1QHgu0Tr8M2A0Uz2p9LrjgAi0IFixYEOkq5Bpra95266236kcffRR0Wc2aNXXXrl2qqrpx40Y9ceKEqqouW7ZMK1SooCdOnNBJkyZphw4dNDk5WY8dO6ZXXXWVzpo1K9fqn9Pyw3cKLNUM/q7a5dYzMw+4SUQqAYhIeRGpiYusO4P9psbw4fbrO6NaGmOyXWpOa2A4eShTp06ladOm+Hw+HnzwQZ5//nlEhJtuuom6devSrFkzWrRoQYsWLbj22lDxziYS7HLrGVDVtSKSGj1XCEgGHuRkZN1yDX5fMjOPAGNcnF0R4AdCR/EZYyIgWE5roPj4eP/0oEGDGDToZEBX6iXIwoULpwlDN3mPdZJnSINHzy0mILJOVUu7n7GkjbyLCZj2L9MQMXzGGGNyl11uNcYYY0KwTtIYY4wJwTpJY4wxJgTrJI0xuapWrVo0a9bMH/YNcOutt9K/f398Ph+1atXC5/MBMHfuXFq1akWzZs1o1aoV8+fP9+9n2bJlNGvWjHr16lkwuMkx9uBOHicitfDen2wa6boYk10WLFhAhQoV/PNffPEFsbGxxMTE8Ne//pWyZcsCUKFCBb766iuqVq3KmjVr6Ny5MwkJCQDcf//9fPDBB1x88cV07dqV2bNnc80110SkPSZ6WScZRSy7NfpEU1vTZ54Go6pMnjzZf8Z44YUX+pc1adKEpKQkjh49yp9//smBAwdo27YtAH369GHGjBnWSZpsZ5dbs4mI3CEiv7qxJN8TkcIi0klEfhaR5SLypYiUduteJCI/ichKt83ZIlJLRBa6dZeLiI0cYqKSiNCpUydatWrF+++/n2bZwoULqVy5MvXr1z9lu6lTp9KyZUuKFStGQkIC1atX9y+rXr26/wzTmOxkZ5LZIF3QebKIvA30BvoDV6tqoogMAh4XkeF471XeqqpLRKQMkISX+dpRVY+ISH1gItA6jGNbdmsUi6a2pr5AP2LECCpWrMjevXt54oknSEpKokWLFhw6dIgPPviANm3anJL3uXnzZoYMGcKIESOIjY1l/fr17N2717/eqlWr2LNnT77JCY2KTNMwREU7M8qss0/YGa4PAduAOPdZDyzFy2BNLVsLfIg3RNaiIPsoC4zHCzqPAw678lrAmnDqYdmt0Sfa2/rCCy/oq6++qqqq//d//6eVKlXSLVu2pFlny5YtWr9+ff3xxx/9Zdu2bdMGDRr45z///HMdMGBA7lQ6G0T795oqP7QTy27NFQKMU294LJ+qNsAbN3JuQFljVb0ng338BdiBN5RWa7zxKY2JKomJiRw8eNA/PWfOHJo29Z5JW7ZsGQ0bNkxzGXXfvn1069aN4cOHc+mll/rLq1SpQpkyZVi8eDGqyqeffkqPHj1ytzGmQLBOMnucEnQOrAIuFZF6rqyUiFyAd5ZZRUQucuVni0gRvDPJ7ap6Am+cycIRaIcxOWrHjh20b9+eFi1a0KZNG7p160aXLl0AmD9/vn/w4lSjR49m06ZNDBs2DJ/Ph8/nY+dObzS6t99+m/79+1OvXj3q1q1rD+2YHGH3JLOBhg467wtMFJFibtUhqrpBRG4F3hKREnj3I68G3gamikgfYDaQmNvtMCan1alTh5UrVwZdNnjwYGJiYtKUDRkyhCFDhgRdv3Xr1qxZsya7q2hMGtZJZhMNHnQOcFGQdZcAbdMVbwSaB8wPcuvGA/aOpDHGRIBdbjXGGGNCsE7SGGOMCcE6SWNMtguWz5pq5MiRiAi7d+8GYO/evfTs2ZN77rmHNm3a+O8zbtmyhSuvvJLGjRvTpEkTRo0alevtMMbuSWYjEYkHWqs3aLIxBVr6fFbwOr45c+ZQo0YNf9nLL7+Mz+fj0Ucf5bzzzuPBBx9k3rx5FClShJEjR9KyZUsOHjxIq1at6NixI40bN87tppgCzM4k8xERsddCTL72l7/8hREjRiAi/rK1a9dy1VVXAdCwYUPi4+PZsWMHVapUoWXLlgCcffbZNGrUyKLnTK6zM8nTJCJ3AI/gvfT/C/BAuuUzgPOB4sAoVX1fRPoBzVX1MbfOvUBjVf1LsP2paoqIHALew3tN5EHgx1B1soDz6JPf2poaYp6azyoi3HfffQwYMICZM2dSrVo1WrRokWabFi1aMG3aNHr06MGvv/7Kf//7X7Zu3UrlypVP7jc+nhUrVnDxxRfnanuMEbUx2LLMZbWOAG7Qk1mti4FhuMutIlJeVf9070IuAa4AjgIrgYZuu5+A+4Djwfanqp+KiOLlvE4OUZfA7NZWz7/xQY62PS+oXAJ2JEW6Frkjv7W1WTVviKtdu3alyWd95JFHePfdd3n11VcpXbo0t912G++99x5ly5YlMTGR0aNHs379eurVq8cff/zBE088Qb169QBISkri0Ucf5Y477uDyyy+PZPOyzaFDhyhdunSkq5Hj8kM7r7zyymWqGjonO6PMOvtkKat1KBAPVHDrDMXrEFcC+4G2rvwDoCfQEFiS0f7csuNA4XDqZdmt0Sca2vrCCy/osGHDtGLFilqzZk2tWbOmFi5cWM8//3zdvn27f70FCxboiRMntGbNmrp//35VVT127Jh26tRJR44cGanq54ho+F7DkR/aSSbZrXa59fSkZrU+naZQpK/7GYN3ebSdqh4WkVi8y64AY4FngN+BjzPan3NEVVOyuf7G5JjExEROnDjB2Wef7c9nff755/1xcuA9/bp06VIqVKjAvn37KFmyJABjx47l8ssvp0yZMqgq99xzD40aNeLxxx+PVHNMAWed5OmZB8wUkX+q6k6X1Xp2wPKywF7XQTYkIF1HVX8RkfOBlpxM2Am6P1X9b+40x5jss2PHDnr27AnA8ePHuf322/35rMGsW7eOu+66iyNHjtC6dWs+/PBDABYtWsT48eP9r5KA9yRs165dc7wNxqSyTvI0aOis1lSzgYEisg7v0unidLuYDPhUdW8m+7NO0uQ7GeWzpoqPj/dPt2vXjg0bNhAbG5smu7V9+/aptzeMiRjrJE+TBs9qrRUwndGQBO2Bf4axP1Q1b9/1NsaYKGbvSeYiESknIhuAJFWdF+n6GGOMyZidSeYiVd0HXBDpehhjjAmPnUkaYzIULIf1zz//pGPHjtSvX5+OHTuyd+9eAF599VX/4MhNmzalcOHC/Pnnnxw5coQ2bdrQokULmjRpwgsvvBDJJhkTNuskjTGZWrBgAXFxcSxduhSA4cOH06FDBzZu3EiHDh0YPnw4AE8++SRxcXHExcXxyiuvcMUVV1C+fHmKFSvG/PnzWblyJXFxccyePZvFi9M/z2ZM3mOdZDoiEiMiX4dYNlZEGrvpZzLZTy0RyZZh0100nTF5xsyZM7nrrrsAuOuuu5gxY8Yp60ycOJFevXoBXkxdavJKcnIyycnJafJbjcmrouaepIgUUdXjOXkMVe0fMPsM8HJOHi+rLLs1+kS6rfHDuwXNYU0NIAc477zz2LFjR5rtDh8+zOzZsxk9erS/LCUlhVatWrFp0yYefPBBy2E1+ULEOskQAeH7gVFAdyAJ6KGqO0SkIvAukDq+zmOqukhEhgJ1gTrAHyLyCPA5UBX4GegItHLH+VNV33DHfgnYqaqhBqgrLSJTgKbAMuAOVVWXnPMEcBNQQkTigN9UtXeI/RQWkQ+AS4AE154kEann2lMRSAFuBnYAM4FzgKLAEFWdGcbvMTC7leeb5ei/E/KEyiW8zqMgiHRbY2NjGTFiRJoc1qSkJI4fP05sbKx/vZSUlDTz8+fPp2HDhqxatSrN/t544w0OHTrEc889R8OGDaldu7Z/2aFDh9LsI5oVlLZGRTszyqzLqQ/QCPgKKOrm3wb6AApc68pG4HUU4HV87d10DWCdnsxHXQaUcPOjgafddBe3vwp47y8ud+WFgH8D54aoWwxeZ13drftzwLFj8QLMAQ5l0sZaeLmrPjc/Ga+zBe8fBT3ddHGgJN4/WMq4sgrAJk4G0Gd4rNSPZbdGn7zW1hdeeEFfffVVveCCC3Tbtm2qqrpt2zZN/9/e9ddfrxMmTAi5nxdffFFfffXVNGV5ra05qaC0NT+0k0yyWyN1T7ID3hneEnc21gHvbPAYkHo/cBknX86/Ghjt1p0FlBGR1JfsZ6lq6jgJ7YFJAKo6G0hNtIkH9ojIhUAnYIWq7smgfr+q6lZVPYEXOF4rg3UzsllV4wLbIyJnA9VUdbqr2xFVPYyX3/qyiKwC/g+oBlQOsk9jck1iYiIHDx70T8+ZM4emTZty3XXXMW7cOADGjRtHjx49/Nvs37+f77//Pk3Zrl272LdvH+CN6jF37lwaNmyYew0x5jRF6nJrqIDwJ1zPDt5lyNT6FcIbReNIuvUBEsM85ligL3Ae8FEm6x4NmA6sR1al30+JDNbtjXf5tZV6w2XFczIU3ZiICJXDetFFF3HLLbfw4YcfUrNmTSZPPjmS2/Tp0+nUqROlSpXyl23fvp277rqLlJQUTpw4wS233EL37t1zvT3GZFWkOsnMAsLTmwM8DLwKICK+gDO0QIuAW4B/iEgnvPt7qabjjfdYFLj9zJtAsogUVdXkrGykqgdFZKuIXK+qM0SkGFAYLxR9p+sgrwRqZkMdjTkjoXJYzz33XObNCx4a1bdvX/r27ZumrHnz5qxYsSInqmhMjorI5VZVXQukBnqvAuYCVTLY5BGgtYisEpG1wMAQ670IdHKvXtwM/A846I55DFgATNbsGXrqfWCViEw4jW3vBB5xbf8J7+x2Al4bV+Pdn/09G+pojDHmDETs6VYNHuhdOmD5FGCKm94N3BpkH0PTFe0HOqvqcRFpB1ykqkcB3OgabfE6z4zqFYv3gE7q/EMB0zEB04OAQRnsJx7v6djU+dcCpjcCVwXZrF2IfVnIuTHGREDUvCfp1AAmuw7xGHAvgAsA+BqY7jooY4wxJlNRlbijqhtV9UJVbaGqF6nqEle+VlXrqOpfU9cVkWYiEpfu80tWjyki5wbZT5yInJudbTMmu6WkpHDhhRf6H6C57LLL/LmrVatW5frrrwe818QeeeQR6tWrR/PmzVm+fDngRdWlru/z+ShevHjQ5B1j8rNoO5MMm6quBnzZsJ894exHRGoBX6tqUxFpDfRR1UfO9PjGnK5Ro0bRqFEjDhw4AMDChQv9y2688Ub/KxzffvstGzduZOPGjfzyyy/cf//9/PLLL1x55ZXExcUBXuB5vXr16NSpU663w5icFFVnkpEgIoWzuo2qLrUO0kTS1q1b+de//kX//v1PWXbgwAHmz5/vP5OcOXMmffr0QURo27Yt+/btY/v27Wm2mTJlCtdccw0lS5bMjeobk2sK7JlkuDKIz3sPL+TgQRG5CrgW7z3In4D7VFVFpBUn38mcE7DPGOAJVe3uovUOpT7Y457M7Q7swkvpqY73isjf3MNOIVl2a/TJibbGD+/GY489xogRI/xBAYFmzJhBhw4dKFOmDAAJCQmcf/75/uXVq1cnISHBn90KMGnSJB5//PFsracxeYF1khkQkUZ4T9Ve6t5ffBvvpf9SwC+p9zhFZK2qDnPT4/E6ua+Aj4GHVPUHEXk1i4fvAmxT1W5uv2VD1NGyW6NYTrT1lVdeITk5mYMHDxIXF8eePXvS5GuOGTOGrl27+sv27NnDihUrOH7cq8fevXtZtmwZhw4d8i9fvnw5xYsXP6OczqjI+QxTQWlrVLQzo8y6gv4BHgK24UXTxQHr8fJijwOFA9a7Ee8sczVekPlgoBzwR8A6zYE1ejIf9ms9mT/7RMB6a/Bi8C4A4oF/AJeFU1/Lbo0+OdHWwYMHa7Vq1bRmzZpauXJlLVGihPbu3VtVVXft2qXly5fXpKQk//oDBgzQzz//3D8fmNuqqvrGG2/ovffee8b1su81+uSHdpJHs1vzi9T4PJ/7NFDv3cwj6gIJRKQ4XkD7TaraDPiArMXJHSftveHiAKq6AWiJ1/H+XUSeP+PWGIN3Jrl161bi4+OZNGkSV111FZ999hng3Vvs3r07xYuf/E/4uuuu49NPP0VVWbx4MWXLlk1zqTVw3Ehjoo11khmbB9wkIpUARKS8iKSPi0v9a7Lbha7fBKCq+4B9ItLeLQ81nFY8XmeIiLQEarvpqsBhVf0ML46vZXY0yJiMTJo06ZQOr2vXrtSpU4d69epx77338vbbb/uXxcfHs2XLFq644orcrqoxucLuSWZAVdeKSGp8XiEgGXgw3Tr73JiRa/Bi8JYELL4b+EhElIAHd9KZCvQRkd/wLtlucOXNgFdF5IQ77v3Z1Cxj/GJiYoiJifHPB7t/JCKMGTMm6Pa1atUiISEhh2pnTOSF1UmKSF1gq6oedU9mNgc+dWdLUU0zic9z6wzBy6JNv+0yoEVA0VOuPBYXfafeMF/BXi6LB747vVobY4zJDuFebp0KpIhIPbxg7/PxBkI2xhhjola4neQJVT0O9ATeUtUnyXjUDmOMMSbfC7eTTBaRXsBdeEHh4I3LaIyJsPQZrL1796ZBgwY0bdqUfv36kZzsDXk6YcIEmjdvTrNmzbjkkkvSjBP5z3/+kyZNmtC0aVN69erFkSNHgh7LmIIm3E7ybrxhnF5S1c0iUhsYn3PVyj9EpJZLyUlf3tc9oZo6P9aNRmJMtkrNYE3Vu3dvfv/9d1avXk1SUhJjx44FoHbt2nz//fesXr2a5557jgEDBgBeos6bb77J0qVLWbNmDSkpKUyaNCkibTEmrwmrk1RvkORBwHI3v1lV/5GTFYsCfQF/J6mq/d3v0ZhsEyyDtWvXrogIIkKbNm3YunUrAJdccgnnnHMOAG3btvWXAxw/fpykpCSOHz/O4cOHqVq1KsaY8J9uvRZ4DS+/tLaI+IBhqnpdDtYtPynsXgO5BC9xZzzQGpggIkl4Z+Hf4iXrLBWRQ8A7QFdgO/AMMAJvPMzHVHWWC04fjpfOUwwYo6rvZVQJy26NPhm1NbMM1uTkZMaPH8+oUaNOWfbhhx9yzTXXAFCtWjWeeOIJatSoQYkSJejUqZON5mGME+57kkOBNpx8bSFOROrkUJ3yo/pAL1W9V0QmAwosxXWK4L1rFqAUMF9VnxSR6cDfgY5AY2AcMAu4B9ivqheJSDFgkYjMUdXNgTuy7NbollFbM8tgfe2116hTpw4pKSlpylesWMFbb73Fm2++SWxsLAcPHmTcuHF89tlnlC5dmqFDh/Lss8/SsWPHHG5dWlGR8xmmgtLWqGhnRpl1ejJPdLH7uSKgbFU420b7By9ndWPA/CC8dyZjgdYB5f554CggbnoY8KybLgTsc9NT8IIF4txnM9Apo7pYdmv0yaitGWWwDh06VHv06KEpKSlptlm5cqXWqVNH169f7y+bPHmy9uvXzz8/btw4vf/++7O3IWGw7zX65Id2kk3Zrb+JyO14lxXri8hbeENCGc/RgOkUMj9DT3ZfDsCJ1O1V9UTAtgI8rCdzY2uraqjUHlMAhcpgHTt2LN999x0TJ06kUKGT/4v/8ccf3HDDDYwfP54LLrjAX16jRg0WL17M4cOHUVXmzZuX5kEgYwqycDvJh4EmeH/MP8cbT/GxHKpTtDgInH0G238H3C8iRQFE5AIRKZUtNTNRbeDAgezYsYN27drh8/kYNmwYAMOGDWPPnj088MAD+Hw+WrduDcDFF1/MTTfdRMuWLWnWrBknTpzwP/lqTEGX6T1J9wDJv1T1SuDZnK9S1PgEeDfgwZ2sGot3KXe5eDc0dwHXZ1flTHQJzGBNHfcxvbFjx/pfB0nvxRdf5MUXX8yp6hmTb2XaSapqioicEJGyqro/NyqVn6hqPNA0YP61gMVTA6ZjAtYpHTA9NN3+SrufJ/Ceen0mO+trjDEmfOE+3XoIWC0ic4HE1EJVfSRHamWMMcbkAeF2ktPcxxhjjCkwwk3cGRfsk9OVMyYrtmzZwpVXXknjxo1p0qSJ/yX6L7/8kiZNmlCoUCGWLl3qXz8+Pp4SJUrg8/nw+XwMHDjQv2zZsmU0a9aMevXq8cgjj3DyYWRjTEESbuLOZrwX5NNQ1agOFBCRWsDXqto0s3XD3F883ruSu0XkJ1W9JDv2azxFihRh5MiRtGzZkoMHD9KqVSs6duxI06ZNmTZtGvfdd98p29StW5e4uLhTyu+//34++OADLr74Yrp27crs2bP9CTXGmIIj3MutrQOmiwM3A+Wzvzr5j4gUUW8YsSyxDjL7ValShSpVvBHczj77bBo1akRCQkKWk2O2b9/OgQMHaNu2LQB9+vRhxowZ1kkaUwCF1Umq6p50RW+IyDLg+eyvUp6TPpe1B14OaxzQHpgoIhvwUnbOAvYAvVV1h4icC0wEqgE/4wUEACAih1S1tIjE4MXXdXflo/ESID4RkeHAdcBxYI6qPpFRRQtydmv88G5p5+PjWbFiBRdffHGG+9q8eTMXXnghZcqU4e9//zuXXXYZCQkJVK9e3b9O9erVSUhIyL4GGGPyjXAvt7YMmC2Ed2YZ7llofpc+l/VGV36WqrYGEJFzgLaqqiLSH3gK+CvwAvCjqg4TkW54eaxhcR1sT6Ch22+5EOtZdiukyYdMSkri0UcfpX///ixfvtxfvm/fPpYtW8ahQ4cAOHbsGJ9//jlly5Zl/fr13HjjjXz88cds3bqVvXv3+ve5atWqU3JRc0tUZF+GydoafaKinRll1unJ3NEFAZ+5wPtAg3C2zc8fMs5lvSKgvBkwB1gNrAdmu/I4oE7Aen8CFdz0IfczBu++Z+o6o/GG2SoCrAQ+Am7A65QzrK9lt6oeO3ZMO3XqpCNHjjxl2RVXXKFLliwJuW3q8m3btmmDBg385Z9//rkOGDDgjOp8uvJD9mV2sbZGn/zQTrIpu/UeVb3SfTqq6gDgWNa643wrVC5rYkD5W8BoVW0G3Id33zZcx0n7lHFxAPXuc7bBCzrvDszOWrULHlXlnnvuoVGjRjz++OOZrr9r1y5SUlIA+M9//sPGjRupU6cOVapUoUyZMixevBhV5dNPP6VHjx45XX1jTB4Ubic5Jcyygqos3v1KgLsCyn8AbgcQkWuAc4Js+1+gsYgUc5dUO7j1SwNlVfUb4C9Ai5ypevRYtGgR48ePZ/78+f7XOr755humT59O9erV+fnnn+nWrRudO3cG4IcffqB58+b4fD5uuukm3n33XcqX955He/vtt+nfvz/16tWjbt269tCOMQVUhvcVRaQhXrB5WRG5IWBRGbJ2thTthgJfisheYD5Q25W/iPdgz294o6b8kX5DVd3i7nWuwRsOa4VbdDYwU0SK4z3wk/mpUQHXvn37kO8z9uzZ85SyG2+8kRtvvDHI2tC6dWvWrFmTrfUzxuQ/mT180wDvUl854NqA8oPAvTlUpzxDM85lDVxvJjAzSPkeIOgQ75o2v/UpvId90muTtRobY4zJThl2kql//EWknar+nEt1MsYYY/KEcF/jWCEiD+JdevVfZlXVfjlSK2OMMSYPCPfBnfHAeUBn4HugOt4lV2My1a9fPypVqkTTpifT/Z588kkaNmxI8+bN6dmzJ/v27QNgwoQJ/odufD4fhQoV8sfGHTt2jAEDBnDnnXfSsGFDpk6dGuRoxhiTfcLtJOup6nNAonrB5t2AjKNMjHH69u3L7Nlp32Dp2LEja9asYdWqVVxwwQW88sorAPTu3Zu4uDji4uIYP348tWvXxufzAfDSSy9RqVIlxo8fz9q1a7niiityuynGmAIm3E4y2f3cJyJN8V55qJQzVcq/ROQTEbkp0vXIay6//HL/qxWpOnXqRJEi3tX+tm3bsnXr1lO2mzhxIrfddpt//qOPPuLpp58GoFChQlSoUCEHa22MMeHfk3zfRa89B8wCSlMwclvzlMzC1PNidmv6TNVgPvroI2699dZTyr/44gtmzvQeGk69HPvcc8/x9ddf06JFC0aPHk3lypWztb7GGBMo3IDzsW7yeyBPDo/lhrWaDSzGCyNfAnyM965iJaA30BUvDu41t80aoLuqxovI40Dqg0hjVfUNt89vgR8JCDhX1aQw6vM83mszJfDekbwP73f3paq2dOvUB75Q1ZYi0gp4He8fILuBvqq6XURiCQhTB0amO06ezm5NzW383//+R2Ji4ik5jp999hn79u2jWrVqaZatXbsWVWX37t3Exsayf/9+tm7dStmyZXn99df55ptvuPPOO3nmmWdyrzEREBXZl2GytkafqGhnRpl1ejJPtDLwIfCtm2+MF1UX8XzVgDrWwot4a4Z3GXkZXu6p4I3cMQPvpf8nArZZ47ZrhZe7Wgqvk/oNuDBgnz63/mTgjgzq8Alwk5suH1A+HrjWTS8I2N/LwMNAUbyOtKIrvxX4yE3HAm+H8zvIy9mtmzdv1iZNmqQp+/jjj7Vt27aamJh4yvqPPfaYvvTSS/75EydOaMmSJTUlJUUXLFigf/zxhzZu3DjH6x1p+SH7MrtYW6NPfmgn2ZTd+gnwHVDVzW8AHgtz29y0WVVXq+oJvI5unvslrMbr8EJpD0xX1URVPQRMAy4L2Gecm16WyX4CXSkiv4jIauAqvNdnAMYCd4tIYbzO8HO80IamwFwRicMLUa8esK8vwjxmvjF79mxGjBjBrFmzKFmyZJplJ06cYPLkyWnuR4oI1157rf9fpfPmzaNx48a5WWVjTAEU7j3JCqo6WUSeBi98W0RScrBepyswjPxEwPwJvLYGDRPPwj5T8C6fZshFyb0NtFYvdm5owLGm4g2hNR9Ypqp7RKQq8Juqtguxy8QQ5flCr169iI2NZffu3VSvXp0XX3yRV155haNHj/oHRG7bti3vvvsu4GWqnn/++dSpk/bK/j/+8Q/uvPNOEhISqFWrFh9//HGut8UYU7CE20kmuvENFUBE2gL7c6xWOSceL2YvdYzM1IzVhUDqIMeCN47jnWdwnNQOcbcLKr8JFwivqkdE5DvgHU6OL7keqJiabCQiRYELVPW3M6hDnjFx4sRTyu65J/TQmjExMSxevPiU8po1a/LDDz8QGxtLTExMdlbRGGOCCreTfBzvqda6IrIIqIj3hz+/mQr0cYHjv+BdNkZVl4vIJ8Cvbr2xqrrCPbiTZaq6T0Q+wLvn+T+8h4gCTcDriOe49Y+5V0feFJGyeN/LG3iXjI0xxkRIZqOA1FDVP1wncgXevTMB1qtqckbb5jY9NYy8b4hloQLHX8d7ujSjfQYNOA9xzCF49xaDaQ98rKopAevHAZcH2WdMRsc0xhiTczI7k5wBtHTTX6hq8HGFTNhEZDpQF+9hHmOMMXlYZk+3SsB0nnw/MhJEZIyIxKX73B3OtqraU1Wbq+runK5nXpCV3Na5c+fSqlUrmjVrRqtWrZg/f75/m2effZbzzz+f0qVLpz+EMcbkmMw6SQ0xXaCp6oOq6kv3sUctg8hKbmuFChX46quvWL16NePGjePOO08+O3Xttdfy66+/YowxuSmzTrKFiBwQkYNAczd9QEQOisiB3KhgfiUifd2rHanzY0WkwL3Yl5Xc1gsvvJCqVb1fWZMmTUhKSuLo0aP+9apUqZKLNTfGmMwHXS6cWxWJQn3xnm7dBqCq/XP6gHkpuzWczFYInds6depUWrZsSbFixbK7asYYE7ZwXwHJVyKd4+pe52gNTBCRJKCd2/YJVV0qIofw3pPsCmwHngFGADWAx1R1lkvkGQ7EAMWAMar6XpBj5cns1sC8xqzmtm7evJkhQ4YwYsSIU7ZJSUmJjjzIMFlbo1NBaWtUtDOjzLr8+iFv5LjG4iXunDKPd3/3Gjc9He99yaJACyDOlQ8AhrjpYsBSoHZG7c6r2a1ZyW3dsmWL1q9fX3/88ceg+ypVqlS+yIPMLtbW6FRQ2pof2kkm2a1ReSbpbFbV1QAuPGCeqqrLUq2FN7JGMP4cV7dtao7rLE4/xzW9Y3hnuuB1yEdVNTmgbuC9z9k8YHzKskB9YPNpHjPPSM1t/f7779Pktu7bt49u3boxfPhwLr300gjW0BhjPOEGnOdHuZHjerr/yEh2/4JJUzf1gtlT9ynAw3ry6dnaqjrnNI8XMb169aJdu3asX7+e6tWr8+GHH/LQQw9x8OBBOnbsiM/nY+DAgQCMHj2aTZs2MWzYMHw+Hz6fj507dwLw1FNPUb16dQ4fPszNN9/M0KFDI9gqY0xBEc1nkpmJJ2dzXA8CZ59B/b4D7heR+e4s8wIgIfUMN7/ISm7rkCFDGDIkeEjRiBEjGDFiBIBltxpjck1B7iRzOsf1E+DdgAd3smos3qXX5SIiwC7g+tPYjzHGmNMUlZ2k5o0c16l4HXGqmIBlpQOmh6bbrrT7eQLvqddnMjqOMcaYnBPN9ySNMcaYM2Kd5Bk6kxzXaBYss/XLL7+kSZMmFCpUiKVLl/rLJ0yY4H9Qx+fzUahQIeLi4gDvnmazZs1o3rw5Xbp0YffuAhF5a4zJI6yTPEOaQY6rCw04YyISLyIVsmNfuSVYZmvTpk2ZNm0al1+edkSw3r17ExcXR1xcHOPHj6d27dr4fD6OHz/Oo48+yoIFC1i1ahXNmzdn9OjRudkMY0wBZ52kyRHBMlsbNWpEgwYNMtxu4sSJ3HbbbcDJoIvExERUlQMHDvizXY0xJjdE5YM7kSAidwCPAGfhPS37gCt/Ce9VkyS8GLsdIlIZeJeTw4/dr6o/icgM4Hy8dzZHqer7WalDXsluDTe3NZgvvviCmTNnAlC0aFHeeecdmjVrRqlSpahfvz5jxoxh4cKF2VVVY4zJkHWS2UBEGgG3Ape6dxrfxsuHLQUsVtVnRWQEcC/wd+BN4HtV7ekyWlOfdu2nqn+KSAlgiYhMVdU9mRw7z2W3pmY1hsps3bdvH8uWLePQobRXo9euXYuqsnv3bmJjYzl+/Dgvv/wy77zzDlWrVuXNN99kwIAB9OzZM//nQYYpKrIvw2RtjT7R0E7rJLNHB7zM1yXeK42UAHbixc997dZZBnR001cBfQBUNQXY78ofEZGebvp8vBi6DDtJd7b5PkCNOvV05OrIf6XxvWO8n/HxlCpV6pQX/8uVK0erVq1o3bp1mvKZM2fSv39///pLlizhnHPOoXfv3gAULlyY4cOHU7p06QITJlCQghOsrdEnGtoZ+b+o0UGAcar6dJpCkScC4ucyjLETkRjgaqCdqh4WkVjCi8rzK1G0MOvP4FJnJJ04cYLJkyenuZRarVo11q5dy65du6hYsSJz586lUaNGEaylMaagsQd3ssc84CYRqQQgIuVFpGYm69/v1i0sImXxAsz3ug6yIdA2pyudk4Jltk6fPp3q1avz888/061bNzp37uxf/4cffuD888+nTp06/rKqVavywgsvcPnll9O8eXPi4uJ45hnLVjDG5B47k8wGqrpWRIYAc0SkEJAMPJjBJo8C74vIPXhnmPfjjQoyUETWAevxxsLMt4JltgL07NkzaHlMTAyLF5/a5IEDB/oD0I0xJrdZJ5lNVPUL4It0xYHxc1OAKW56B964luldE2LftbKnlsYYY7LCLrcaY4wxIVgnaYwxxoRgnaTJdlnJbd2zZw9XXnklpUuX5qGHHvKXHzx4ME2ea4UKFXjsscdysxnGGGOdpMl+WcltLV68OH/729947bW0I4+dffbZ/jzXuLg4atasyQ033JDjdTfGmEDWSUZQqOByEblORAa76etFpHHu1+70ZSW3tVSpUrRv357ixUO/ErphwwZ27tzJZZddlu11NcaYjNjTrVkkXqSOuEGRc4SqzgJmudnr8VJ71ma2XaSzW88kszUjkyZN4tZbb8WlGRljTK7JsU5SRGrhvfu3GLgEWAJ8DLwIVMLLNu0KHFLV19w2a4DuqhovIo8D/dzuxqrqG26f3wI/un0m4IWGJ4WoQ11gDFAROAzcq6q/i8gnwAGgNXAe8JR7RQMReRK4BSgGTFfVF9xxv8MLLm8FdBWRPsAdwC5gC17s3HTgS1Vt6fZVH/gidT6Eh0XkWqAocLOrX19Xt8+B64Ar3HuYN6rqv9O1Mc9ktwZmNGY1t/X3338nISEhaM7jRx99xNNPP+1fFg15kOGytkangtLWaGhnTp9J1gNuxuvslgC3A+3x/vA/A8QF20hEWgF3AxfjRb79IiLfA3vx8kx7qeq9IjIZuBH4LMTx3wcGqupGEbkYeBsvNxWgiqtLQ7yztiki0sntv4077iwRuRz4w5XfpaqLReQid9wWeJ3bcmCZqv5bRPaLiE9V41wbPs7kd7RbVVuKyAPAE0D/1AVuZJBZwNepnXh6eSm7NTWzFbKe2xofH8+hQ4dOWX/lypWcddZZ3Hffff6yaMiDDJe1NToVlLZGQztz+i/qZlVdDSAivwHzVFVFZDVQixCdJF7nNV1VE92204DL8Dqzza4DAu/srVawHYhIabyzzS8DLtMVC1hlhrtkutYNXQXQyX1WuPnSeJ3jH8B/VTU1EuZSYKaqHgGOiMhXAfsdC9ztzoRvxetwMzItoC1n9GRKfs5uDWXixIn06tUr0tUwxhRQOd1JHg2YPhEwf8Id+zhpHx4KJ9A7cJ8peCNuBFMI2KeqvjD2IwE/X1HV9wJXdJdbE8OoG8BU4AVgPt7ZZYajeATUI8MA9PykV69exMbGsnv3bqpXr86LL75I+fLlefjhh9m1axfdunXD5/Px3XffAVCrVi0OHDjAsWPHmDFjBnPmzKFxY+9ZpcmTJ/PNN99EsjnGmAIs0n+U4/EGJEZEWgK1XflC4BMRGY7XcfUE7szKjlX1gIhsFpGbVfVL98BNc1VdmcFm3wF/E5EJqnpIRKrh5bCmtwh4T0Rewfsddsdd8lTVIyLyHfAOcE9W6hzCQeDsbNhPrslqbmt8fHzIff3nP//JjioZY8xpifQrIFOB8u5S7EPABgBVXQ58AvyK97DMWFVdEWonGegN3CMiK4HfCJ6X6qeqc/AelvnZXRKeQpAOSlWX4F36XYX3INFqTo4JCTAB72x5zmnUOb1JwJMissI9iGSMMSaX5NiZpKrGA00D5vuGWNYpxPavA69nss/XyICqbga6BCnvm24+MIh8FDAqyO6appt/TVWHikhJ4Ae8e4qp2gMfuwGVM6pfrYDppUCMm/4E7x8JqOoiIF+9J2mMMdEi0pdb87P33Uv+xfEGXF4OICLTgbqcfIrWGGNMPhXpy63ZQkTGiEhcus/dOXlMVb1dVX2q2lBVXwko76mqzVV1d0D9pgepX+fge87fspLbCvDKK69Qr149GjRo4H+QB7yHeZo1a4bP5zvldRFjjMktUXEmqaoZDXAccaoa/ImVKNS3b18eeugh+vTp4y9LzW0NfNcRYO3atUyaNInffvuNbdu2cfXVV7NhwwYKFy4MwIIFC6hQ4ZTUPmOMyTVRcSZ5JkQkVkROOVURkdYi8qabjhGRS3K/dv66DBWRJyJ1/KzISm7rzJkzue222yhWrBi1a9emXr16/Prrr7lVVWOMyVRUnEnmBPcgTeq1wRjgEPBTxCoUhvyW3ZqQkEDbtm3989WrVychIQEAEaFTp06ICPfddx8DBgzI1roaY0w4CkwnGSr31S2+WUTeBsoB96jqQhGJwYuJewgYCKSIyB3Aw3h5ry/gBQDsV9W04z+dPGZhYDheJ1sMGJMaVCAig/CyX08A36rqYBG5Fy+H9SxgE3Cnqh7OpF15Lrs13NzWhIQE1q1b519v+/bt/Pbbb1SoUIERI0ZQsWJF9u7dyxNPPEFSUhItWrQAoiMPMlzW1uhUUNoaDe0sMJ2kEyz3FaCIqrYRka54nd/VqRu4sPV3SRvEvhrorKoJIlIug+Pdg9eJXiQixYBFIjIHLy+2B3Cxqh4WkdTrk9NU9QN3jL+77d/KqEF5Mbs13NzWn3/+GcC/3iuvvEKnTp1o165dmu1WrlxJcnKyf71oyIMMl7U1OhWUtkZDOwtaJxkq93VakLKMLMJLBJocsG0wnYDmInKTmy+L11Ffjfce5WEAVf3TLW/qOsdyeLmx35EF+S279brrruP222/n8ccfZ9u2bWzcuJE2bdqQmJjIiRMnOPvss0lMTGTOnDk8//zzka6uMaYAKmidZKjc1yzlp6rqQDeqSDdgmYi0CpHRKsDDqpqms8vg9Y9PgOtVdaUbLisms7rkNVnJbW3SpAm33HILjRs3pkiRIowZM4bChQuzY8cOf4Td8ePHuf322+nS5ZRMCGOMyXEFrZM8XQeBMqkzIlJXVX/BG8LrGuB8IFgn+R1wv4jMV9VkEbkA717oXOB5lxF7WETKu7PJs4HtIlIUL1IvIYfble2ymtv67LPP8uyzz6Ypq1OnDitXZhSxa4wxucM6yfB8hTfeZA+8B3f+4gZUFmAeEOov+li8y7fLXcD6Lrwzxdki4gOWisgx4Bu88TWfw8uq3eV+5qtgc2OMiTYFppMMJ/fVpeTUctOxQKyb3gA0D1h1YZjHPIHX+T0TZNlwvCdfA8vewRs9JP26Q8M5njHGmOxV4MMEjDHGmFCsk8wGItI5SDbr9EjXK6eNGjWKpk2b0qRJE9544w3g9HJajTEmryowl1tzknt6tUD91V+zZg0ffPABv/76K2eddRZdunShe/fup53TaowxeZGdSeYCEekrIqNDLPtGRMq5zwO5XbfTtW7dOi6++GJKlixJkSJFuOKKK5g2bZrltBpjooqdSUaYqnYFf2zeA8Dbp7uv3MpujR/ejaZNm/Lss8+yZ88eSpQowTfffJPhkFYZ5bQaY0xeFTVnkiJSS0R+F5FPRGSDiEwQkatFZJGIbBSRNulH0xCRNa5zQkQed/NrROSxgH2uE5EPROQ3EZkjIiWC18A/osg/RORXV4fLAhZXFZHZri4jAraJF5EKeE+61nX3M18VkSoi8oObX5NuXxHXqFEjBg0aRKdOnejSpQs+n88unRpjok60nUnWA24G+gFLgNuB9sB1eK9hxAXbSERaAXcDF+O9+/iLiHwP7CV43utnGdQhVA6sD7gQL91nvYi8papbArYbDDRVVZ+r01+B71T1JReUXjJE3XM94Dw1sLhu3bqMHDkSgA8++ICKFSv6l6UPMz969Cjff/891atXB2DVqlW0bNnytMKPoyE0OVzW1uhUUNoaDe2Mtk5ys6quBhCR34B5qqoukLwWITpJvI50uqomum2nAZcBswid9xpKqBzYeaq63+1/LVAT2EJoS4CPXPrOjIA6pBEYcN6gQQN9uHePYKvliJ07d1KpUiX++OMPli1bxuLFiylXrhxwaph5xYoVuf322xk9ejTbtm1jz549DBw48LTOPqMhNDlc1tboVFDaGg3tjLZOMjCb9UTA/Am8th4n7SXm4lncZ2Dea2brp8+BTb+fDH/3qvqDiFyOlw/7iYi8rqqfhlHfXHPjjTeyZ88eihYtypgxYyhXrhzTp0/PUk6rMcbkZdHWSWYmHugOICItgdqufCFeRzQc73JrT+DOXK7bQQJi6ESkJrBVVT9ww2y1BPJUJ7lw4anBQz179sxSTqsxxuRlBa2TnAr0cZdifwE2AKjqchH5BEh9J2Gsqq5IfagnN6jqHveQ0Rq8waHXAE+KSDJwCOiTW3UxxhjjiZpOMkg2a98QyzqF2P514PVM9nlK3mu69WMCpgNzYD/BGwYrdVn3gOlaAdO3p9vluIyOZ4wxJmdFzSsgxhhjTHazTvI0iMiYIFmtd0e6Xtlp37593HTTTTRs2JBGjRrx888/Wy6rMabAiZrLrblJVR+MdB1y2qOPPkqXLl2YMmUKx44d4/Dhw5QrV85yWY0xBYp1khHiBmEWN+ZknrJ//35++OEHPvnkEwDOOusszjrrLP87kOmFymVt165d7lXaGGNyQJ7vJN0TprOBxcAleC/Zfwy8CFQCegNdgUOpD9a4J0S7q2q8iDyOl8AD3lOrb7h9fgv86PaZAPRQ1aQQdagLjAEqAoeBe1X1d/dE7AGgNXAe8JSqTnHbPAncAhTDCyp4wR33O7wna1sBXUWkD3AHsAsvXGAZMB34UlVbun3VB75InQ8lu7JbZ9xWjYoVK3L33XezcuVKWrVqxahRoyhVqlTQ9S2X1RgTrfLLPcl6wEigofukxs09gRc3F1S6uLm2wL0icqFbXB8Yo6pNgH14cXOhvA88rKqt3DEDQ8iruLp0x8tfRUQ6uf23wYuja+WCAVKP+7Y7biV33BbANXidLar6b2C/iPjcNnfj/cMgVxw/fpzly5dz//33s2LFCkqVKsXw4cNz6/DGGJNn5PkzSSdicXMiUhrvbPNL7wop4J0dpprhLpmuFZHKrqyT+6xw86XxOsc/gP+q6mJXfikwU1WPAEdE5KuA/Y4F7nZnwrfidbjB6pft2a1//PEHFSpUICkpidjYWOrWrcvnn39Ohw4dgJzNZQ1HNORBhsvaGp0KSlujoZ35pZOMZNxcIWBfavB4JvuRgJ+vqOp7gSu6y62JYdQNvOCDF4D5wDJV3RNspZzKbv3nP/9JlSpVaNCgAbGxsVx22WX+DMaczGUNRzTkQYbL2hqdCkpbo6Gd+eVya2bi8WLbgsXNXS8iJUWkFF7c3KlZahlQ1QPAZhG52e1fRKRFJpt9B/RzZ6GISDURqRRkvUXAtSJS3K0bGDJwxO3nHXLxUmuqt956i969e9O8eXPi4uJ45plnmD59OtWrV+fnn3+mW7dudO7cGSBNLmuXLl0sl9UYEzXyy5lkZnI6bq438I6IDAGKApOAlaFWVtU5ItII+Nldoj2E93BOSrr1lojILGAVsANYDewPWGUCXsc+J4v1PWM+n++UdyEtl9UYU9Dk+U4yj8TNbQa6BCnvm26+dMD0KGBUkN01TTf/mqoOFZGSwA9490dTtQc+VtUUjDHG5Lo830kWAO+LSGO8+6jjVHU5gIhMB+oCV0WycsYYU5BZJxlARMbgPXEaaJSq5tg9wSCh5qnlwa9rGmOMyTXWSQYoCHFzxhhjwhctT7eaLDpy5Aht2rShRYsWNGnShBdeeAGAyy67DJ/Ph8/no2rVqlx//fX+bWJjY/H5fDRp0oQrrrgiQjU3xpjcU+DPJEUkFnhCVZemK28N9FHVR0QkBjimqj9lcd99gdaq+tAZ1jHG1bF7JquGrVixYsyfP5/SpUuTnJxM+/btueaaa1i48OQbMjfeeCM9enjvXe7bt48HHniA2bNnU6NGDXbu3JldVTHGmDzLziRDUNWlqvqIm43BS92JGiJC6dLew7jJyckkJycTkCjEgQMHmD9/vv9M8vPPP+eGG26gRo0aAFSqFOy1T2OMiS4F5kwyVKi5W3yziLwNlAPuUdWFqWdvwEPAQCBFRO4AHsYLM38B773H/ap6OaFVFZHZeE+qTlfVp1x9ugAvA4WB3araQUTa4L02UhxIAu5W1fXhtjErAefxw7uRkpJCq1at2LRpEw8++CAXX3yxf/mMGTPo0KEDZcqUAWDDhg0kJycTExPDwYMHefTRR+nTp0+4VTPGmHypwHSSTn2gl6reKyKTORlqXkRV24hIV7zO7+rUDdxIIu+SdpSR1UBnVU0QkXKZHNMHXIgXX7deRN4CjgAfAJer6mYRKe/W/R24TFWPi8jVeJ1oRsHrp53dmpqn+MYbb3Do0CGee+45GjZsSO3aXljRmDFj6Nq1q3+9//73v6xfv56RI0dy7NgxHnzwQUSE888/P6zjZadoyIMMl7U1OhWUtkZDOwtaJxkq1HxakLKMLAI+cR3ttEzWnaeq+wFEZC1QEzgH+MGFFKCqf7p1ywLj3NBYipfuk6Hsym5dvnw5e/bs4e6772b37t1s2rSJQYMGUby4F4O7ePFimjdvzjXXXAPArFmzKF68eERyGaMhDzJc1tboVFDaGg3tLGj3JNOHmhdJVx5YFpKqDgSGAOcDy0Tk3NM4ZjB/AxaoalPgWsILaj8tu3btYt++fQAkJSUxd+5cGjZsCMCUKVPo3r27v4ME6NGjBz/++CPHjx/n8OHD/PLLLzRq1CinqmeMMXlCQTuTPF0HgTKpMyJSV1V/AX4RkWvwOsugo3SEsBh4W0Rqp15udWeTZfHulQL0zZ6qB7d9+3buuusuUlJSOHHiBLfccgvdu3sPz06aNInBgwenWb9Ro0Z06dKF5s2bU6hQIfr370/TpukT9owxJrpYJxmer4ApItID78Gdv7hLogLMI4Ow82BUdZe7lzhNRAoBO4GOwAi8y61DgPCewDlNzZs3Z8WKFUGXhbqH8OSTT/Lkk0/mYK2MMSZvKTCdZDih5qq6G3dPUlVjgVg3vQFoHrBqWMNtqeonwCcB84FDYX2L97Rt4Po/AxcEFA1JXxdjjDG5p6DdkzTGGGPCVmDOJHOSiHQG/pGueLOFlBtjTP5mnWQ2UNXvgO8iXQ9jjDHZyy63GmOMMSFYJ2mMMcaEYJ2kMcYYE4KoaqTrYLKJiBwEwg5Ez8cqALsjXYlcYm2NTgWlrfmhnTVVtWKohfbgTnRZr6qtI12JnCYiSwtCO8HaGq0KSlujoZ12udUYY4wJwTpJY4wxJgTrJKPL+5GuQC4pKO0Ea2u0KihtzffttAd3jDHGmBDsTNIYY4wJwTpJY4wxJgTrJKOAiHQRkfUisklEBme+Rd4jIueLyAIRWSsiv4nIo668vIjMFZGN7uc5rlxE5E3X5lUi0jJgX3e59TeKyF2RalNGRKSwiKwQka/dfG0R+cW15wsROcuVF3Pzm9zyWgH7eNqVr3ch+3mOiJQTkSki8ruIrBORdlH8nf7F/be7RkQmikjxaPleReQjEdkpImsCyrLtexSRViKy2m3zpohI7rYwA6pqn3z8AQoD/wbqAGfhDQDdONL1Oo12VAFauumzgQ1AY7yBqAe78sHAP9x0V7zxOAVoC/ziyssD/3E/z3HT50S6fUHa+zjwOfC1m58M3Oam3wXud9MPAO+66duAL9x0Y/ddFwNqu/8GCke6XUHaOQ7o76bPAspF43cKVAM2AyUCvs++0fK9ApcDLYE1AWXZ9j0Cv7p1xW17TaTbnPqxM8n8rw2wSVX/o6rHgElAjwjXKctUdbuqLnfTB4F1eH94euD9ocX9vN5N9wA+Vc9ioJyIVAE6A3NV9U9V3QvMBbrkXksyJyLVgW7AWDcvwFXAFLdK+namtn8K0MGt3wOYpKpHVXUzsAnvv4U8Q0TK4v1x/RBAVY+p6j6i8Dt1igAlRKQIUBLYTpR8r6r6A/BnuuJs+R7dsjKquli9HvPTgH1FnHWS+V81YEvA/FZXlm+5S08XAr8AlVV1u1v0P6Cymw7V7vzw+3gDeAo44ebPBfap6nE3H1hnf3vc8v1u/fzQztrALuBjd2l5rIiUIgq/U1VNAF4D/sDrHPcDy4jO7zVVdn2P1dx0+vI8wTpJk6eISGlgKvCYqh4IXOb+lZmv31kSke7ATlVdFum65IIieJfo3lHVC4FEvMtyftHwnQK4+3E98P5hUBUoRd48280R0fI9BmOdZP6XAJwfMF/dleU7IlIUr4OcoKrTXPEOdzkG93OnKw/V7rz++7gUuE5E4vEujV8FjMK7JJWapRxYZ3973PKywB7yfjvBOyPYqqq/uPkpeJ1mtH2nAFcDm1V1l6omA9Pwvuto/F5TZdf3mOCm05fnCdZJ5n9LgPruKbqz8B4CmBXhOmWZux/zIbBOVV8PWDQLSH0K7i5gZkB5H/ckXVtgv7v08x3QSUTOcf+67+TK8gRVfVpVq6tqLbzvar6q9gYWADe51dK3M7X9N7n11ZXf5p6SrA3Ux3v4Ic9Q1f8BW0SkgSvqAKwlyr5T5w+grYiUdP8tp7Y16r7XANnyPbplB0Skrfvd9QnYV+RF+skh+5z5B+9psg14T8I9G+n6nGYb2uNdrlkFxLlPV7z7NPOAjcD/AeXd+gKMcW1eDbQO2Fc/vAceNgF3R7ptGbQ5hpNPt9bB+2O4CfgSKObKi7v5TW55nYDtn3XtX08eehowXRt9wFL3vc7Ae6oxKr9T4EXgd2ANMB7vCdWo+F6BiXj3WpPxrhDck53fI9Da/d7+DYzGpcHlhY/F0hljjDEh2OVWY4wxJgTrJI0xxpgQrJM0xhhjQrBO0hhjjAnBOkljjDEmhCKZr2KMKchEJAXvUf5U16tqfISqY0yusldAjDEZEpFDqlo6F49XRE/mnRoTUXa51RhzRkSkioj8ICJxbizFy1x5FxFZLiIrRWSeKysvIjPcOIOLRaS5Kx8qIuNFZBEwXkQqishUEVniPpdGsImmALPLrcaYzJQQkTg3vVlVe6ZbfjtevNhLIlIYKCkiFYEPgMtVdbOIlHfrvgisUNXrReQqvGGRfG5ZY6C9qiaJyOfAP1X1RxGpgRdp1ijHWmhMCNZJGmMyk6SqvgyWLwE+cgH1M1Q1TkRigB/UGxMRVU0di7A9cKMrmy8i54pIGbdslqomuemrgcYBA9SXEZHSqnoouxplTDiskzTGnBFV/UFELscbSPoTEXkd2Hsau0oMmC4EtFXVI9lRR2NOl92TNMacERGpCexQ1Q+AsXjDYS0GLncjWRBwuXUh0NuVxQC7Nd24oc4c4OGAY/hyqPrGZMjOJI0xZyoGeFJEkoFDQB9V3SUiA4BpIlIIb6zBjsBQvEuzq4DDnBxqKb1HgDFuvSLAD8DAHG2FMUHYKyDGGGNMCHa51RhjjAnBOkljjDEmBOskjTHGmBCskzTGGGNCsE7SGGOMCcE6SWOMMSYE6ySNMcaYEP4fpJW4sKZbkIkAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "xgb.plot_importance(xg_reg)\n",
    "plt.rcParams['figure.figsize'] = [5, 5]\n",
    "#plt.savefig('aaa')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "#from sklearn.model_selection import GridSearchCV\n",
    "#from sklearn.metrics import r2_score\n",
    "#from sklearn.metrics import make_scorer\n",
    "#ftwo_scorer = make_scorer(r2_score, beta=2)\n",
    "#type(ftwo_scorer)\n",
    "#param_grid = dict(epochs=[10, 20, 30])\n",
    "#grid = GridSearchCV(estimator=model, param_grid=param_grid, scoring=ftwo_scorer, n_jobs=-1, cv=3)\n",
    "#grid_result = grid.fit(x_train, y_train)\n",
    "\n",
    "\n",
    "\n",
    "#from tensorflow.keras.wrappers.scikit_learn import KerasRegressor\n",
    "#estimator = KerasRegressor(build_fn=model, nb_epoch=100, batch_size=100, verbose=False)\n",
    "# Create a KerasClassifier object\n",
    "#model = KerasRegressor(build_fn = model,\n",
    "#                       verbose = 0)\n",
    "#from sklearn.model_selection import RandomizedSearchCV\n",
    "#from sklearn.metrics import r2_score\n",
    "#from sklearn.metrics import make_scorer\n",
    "#ftwo_scorer = make_scorer(r2_score, beta=2)\n",
    "#type(ftwo_scorer)\n",
    "# Define the parameters to try out\n",
    "#params = {'activataion': [\"relu\", \"tanh\"],\n",
    "#          'batch_size': [16, 32, 64], \n",
    "#          'epochs': [50, 100],\n",
    "#          'learning_rate': [0.01, 0.001, 0.0001]}\n",
    "#random_search = RandomizedSearchCV(model,\n",
    "#                                   param_distributions = params,\n",
    "#                                   scoring = ftwo_scorer,\n",
    "#                                   cv = KFold(10))\n",
    "#random_search_results = random_search.fit(x_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "#from sklearn.model_selection import GridSearchCV\n",
    "#from sklearn.metrics import r2_score\n",
    "#from sklearn.metrics import make_scorer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "#ftwo_scorer = make_scorer(r2_score)\n",
    "#type(ftwo_scorer)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "#param_grid = {'epochs': [50, 100, 150, 200],\n",
    "#          'max_depth': [6, 10, 15, 20, 25, 30, 35, 40],\n",
    "#          'learning_rate': [0.001, 0.01, 0.1, 0.2, 0,3],\n",
    "#          'colsample_bytree': [0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1.0]}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "#scoring_fit='neg_mean_squared_error'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "#grid = GridSearchCV(estimator=xg_reg, param_grid=param_grid, scoring=scoring_fit, n_jobs=-1, cv=3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "#grid_result = grid.fit(x_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "#pred = grid_result.predict(x_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "#xg_reg.base_score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "#pred = pd.Series(pred, name='pred', dtype='float32')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "#pred.plot(kind='hist')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "#best_params2 = grid.best_params_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "#print(\"Best params2: \")\n",
    "#for param_name in sorted(best_params2.keys()):\n",
    "#    print('%s: %r' % (param_name, best_params2[param_name]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "#from sklearn.model_selection import RandomizedSearchCV\n",
    "#from sklearn.metrics import r2_score\n",
    "#from sklearn.metrics import make_scorer\n",
    "#ftwo_scorer = make_scorer(r2_score)\n",
    "#type(ftwo_scorer)\n",
    "#Define the parameters to try out\n",
    "#params = {'epochs': [50, 100, 150, 200],\n",
    "#          'max_depth': [6, 10, 15, 20, 25, 30, 35, 40],\n",
    "#          'learning_rate': [0.001, 0.01, 0.1, 0.2, 0,3],\n",
    "#          'colsample_bytree': [0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1.0]}\n",
    "#random_search = RandomizedSearchCV(xg_reg,\n",
    "#                                   param_distributions = params,\n",
    "#                                   scoring = ftwo_scorer,\n",
    "#                                   cv = KFold(10))\n",
    "#random_search_results = random_search.fit(x_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "#pred2 = random_search_results.predict(x_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "#pred2 = pd.Series(pred2, name='pred2', dtype='float32')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "#pred2.plot(kind='hist')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "#best_params = random_search.best_params_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "#print(\"Best params: \")\n",
    "#for param_name in sorted(best_params.keys()):\n",
    "#    print('%s: %r' % (param_name, best_params[param_name]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "#import matplotlib.pyplot as plt\n",
    "#xgb.plot_tree(xg_reg,num_trees=0)\n",
    "#plt.rcParams['figure.figsize'] = [50, 10]\n",
    "#plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "#regr = DecisionTreeRegressor(max_depth=30)\n",
    "#history = regr.fit(x_train, y_train)\n",
    "\n",
    "#y_predict = regr.predict(x_test)\n",
    "\n",
    "#y_predict = pd.Series(y_predict, name='y_predict', dtype='float32')\n",
    "\n",
    "#y_predict.plot(kind='hist')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "#sklearn.tree.plot_tree(regr)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "#plt.plot(y_predict, color=\"yellowgreen\", label=\"max_depth=5\", linewidth=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "file5 = uproot.open(\"./data/gun_k0L_5gev_FTFP_BERT_5000evt_ILD_l5_v02steel.root\")\n",
    "file10 = uproot.open(\"./data/gun_k0L_10gev_FTFP_BERT_5000evt_ILD_l5_v02steel.root\")\n",
    "file15 = uproot.open(\"./data/gun_k0L_15gev_FTFP_BERT_5000evt_ILD_l5_v02steel.root\")\n",
    "file20 = uproot.open(\"./data/gun_k0L_20gev_FTFP_BERT_5000evt_ILD_l5_v02steel.root\")\n",
    "file25 = uproot.open(\"./data/gun_k0L_25gev_FTFP_BERT_5000evt_ILD_l5_v02steel.root\")\n",
    "file30 = uproot.open(\"./data/gun_k0L_30gev_FTFP_BERT_5000evt_ILD_l5_v02steel.root\")\n",
    "file40 = uproot.open(\"./data/gun_k0L_40gev_FTFP_BERT_5000evt_ILD_l5_v02steel.root\")\n",
    "file50 = uproot.open(\"./data/gun_k0L_50gev_FTFP_BERT_5000evt_ILD_l5_v02steel.root\")\n",
    "file60 = uproot.open(\"./data/gun_k0L_60gev_FTFP_BERT_5000evt_ILD_l5_v02steel.root\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "file1 = uproot.open(\"./data/data_other/gun_k0L_1gev_FTFP_BERT_5000evt_ILD_l5_v02steel.root\")\n",
    "file2 = uproot.open(\"./data/data_other/gun_k0L_2gev_FTFP_BERT_5000evt_ILD_l5_v02steel.root\")\n",
    "file3 = uproot.open(\"./data/data_other/gun_k0L_3gev_FTFP_BERT_5000evt_ILD_l5_v02steel.root\")\n",
    "file4 = uproot.open(\"./data/data_other/gun_k0L_4gev_FTFP_BERT_5000evt_ILD_l5_v02steel.root\")\n",
    "file70 = uproot.open(\"./data/data_other/gun_k0L_70gev_FTFP_BERT_5000evt_ILD_l5_v02steel.root\")\n",
    "file80 = uproot.open(\"./data/data_other/gun_k0L_80gev_FTFP_BERT_5000evt_ILD_l5_v02steel.root\")\n",
    "file90 = uproot.open(\"./data/data_other/gun_k0L_90gev_FTFP_BERT_5000evt_ILD_l5_v02steel.root\")\n",
    "\n",
    "file7 = uproot.open(\"./data/data_other/gun_k0L_7gev_FTFP_BERT_5000evt_ILD_l5_v02steel.root\")\n",
    "file45 = uproot.open(\"./data/data_other/gun_k0L_45gev_FTFP_BERT_5000evt_ILD_l5_v02steel.root\")\n",
    "\n",
    "file120 = uproot.open(\"./data/data_other/gun_k0L_120gev_FTFP_BERT_5000evt_ILD_l5_v02steel.root\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [],
   "source": [
    "file6 = uproot.open(\"./data/data_other/gun_k0L_6gev_FTFP_BERT_5000evt_ILD_l5_v02steel.root\")\n",
    "file8 = uproot.open(\"./data/data_other/gun_k0L_8gev_FTFP_BERT_5000evt_ILD_l5_v02steel.root\")\n",
    "file9 = uproot.open(\"./data/data_other/gun_k0L_9gev_FTFP_BERT_5000evt_ILD_l5_v02steel.root\")\n",
    "file12 = uproot.open(\"./data/data_other/gun_k0L_12gev_FTFP_BERT_5000evt_ILD_l5_v02steel.root\")\n",
    "file14 = uproot.open(\"./data/data_other/gun_k0L_14gev_FTFP_BERT_5000evt_ILD_l5_v02steel.root\")\n",
    "file16 = uproot.open(\"./data/data_other/gun_k0L_16gev_FTFP_BERT_5000evt_ILD_l5_v02steel.root\")\n",
    "file18 = uproot.open(\"./data/data_other/gun_k0L_18gev_FTFP_BERT_5000evt_ILD_l5_v02steel.root\")\n",
    "file35 = uproot.open(\"./data/data_other/gun_k0L_35gev_FTFP_BERT_5000evt_ILD_l5_v02steel.root\")\n",
    "file55 = uproot.open(\"./data/data_other/gun_k0L_55gev_FTFP_BERT_5000evt_ILD_l5_v02steel.root\")\n",
    "file65 = uproot.open(\"./data/data_other/gun_k0L_65gev_FTFP_BERT_5000evt_ILD_l5_v02steel.root\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [],
   "source": [
    "f5 = uproot.pandas.iterate(\"./data/gun_k0L_5gev_FTFP_BERT_5000evt_ILD_l5_v02steel.root\", fTreeName, branches)\n",
    "file5 = pd.concat(list(f5))\n",
    "\n",
    "f10 = uproot.pandas.iterate(\"./data/gun_k0L_10gev_FTFP_BERT_5000evt_ILD_l5_v02steel.root\", fTreeName, branches)\n",
    "file10 = pd.concat(list(f10))\n",
    "\n",
    "f15 = uproot.pandas.iterate(\"./data/gun_k0L_15gev_FTFP_BERT_5000evt_ILD_l5_v02steel.root\", fTreeName, branches)\n",
    "file15 = pd.concat(list(f15))\n",
    "\n",
    "f20 = uproot.pandas.iterate(\"./data/gun_k0L_20gev_FTFP_BERT_5000evt_ILD_l5_v02steel.root\", fTreeName, branches)\n",
    "file20 = pd.concat(list(f20))\n",
    "\n",
    "f25 = uproot.pandas.iterate(\"./data/gun_k0L_25gev_FTFP_BERT_5000evt_ILD_l5_v02steel.root\", fTreeName, branches)\n",
    "file25 = pd.concat(list(f25))\n",
    "\n",
    "f30 = uproot.pandas.iterate(\"./data/gun_k0L_30gev_FTFP_BERT_5000evt_ILD_l5_v02steel.root\", fTreeName, branches)\n",
    "file30 = pd.concat(list(f30))\n",
    "\n",
    "f40 = uproot.pandas.iterate(\"./data/gun_k0L_40gev_FTFP_BERT_5000evt_ILD_l5_v02steel.root\", fTreeName, branches)\n",
    "file40 = pd.concat(list(f40))\n",
    "\n",
    "f50 = uproot.pandas.iterate(\"./data/gun_k0L_50gev_FTFP_BERT_5000evt_ILD_l5_v02steel.root\", fTreeName, branches)\n",
    "file50 = pd.concat(list(f50))\n",
    "\n",
    "f60 = uproot.pandas.iterate(\"./data/gun_k0L_60gev_FTFP_BERT_5000evt_ILD_l5_v02steel.root\", fTreeName, branches)\n",
    "file60 = pd.concat(list(f60))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [],
   "source": [
    "f1 = uproot.pandas.iterate(\"./data/data_other/gun_k0L_1gev_FTFP_BERT_5000evt_ILD_l5_v02steel.root\", fTreeName, branches)\n",
    "file1 = pd.concat(list(f1))\n",
    "\n",
    "f2 = uproot.pandas.iterate(\"./data/data_other/gun_k0L_2gev_FTFP_BERT_5000evt_ILD_l5_v02steel.root\", fTreeName, branches)\n",
    "file2 = pd.concat(list(f2))\n",
    "\n",
    "f3 = uproot.pandas.iterate(\"./data/data_other/gun_k0L_3gev_FTFP_BERT_5000evt_ILD_l5_v02steel.root\", fTreeName, branches)\n",
    "file3 = pd.concat(list(f3))\n",
    "\n",
    "f4 = uproot.pandas.iterate(\"./data/data_other/gun_k0L_4gev_FTFP_BERT_5000evt_ILD_l5_v02steel.root\", fTreeName, branches)\n",
    "file4 = pd.concat(list(f4))\n",
    "\n",
    "f70 = uproot.pandas.iterate(\"./data/data_other/gun_k0L_70gev_FTFP_BERT_5000evt_ILD_l5_v02steel.root\", fTreeName, branches)\n",
    "file70 = pd.concat(list(f70))\n",
    "\n",
    "f80 = uproot.pandas.iterate(\"./data/data_other/gun_k0L_80gev_FTFP_BERT_5000evt_ILD_l5_v02steel.root\", fTreeName, branches)\n",
    "file80 = pd.concat(list(f80))\n",
    "\n",
    "f90 = uproot.pandas.iterate(\"./data/data_other/gun_k0L_90gev_FTFP_BERT_5000evt_ILD_l5_v02steel.root\", fTreeName, branches)\n",
    "file90 = pd.concat(list(f90))\n",
    "\n",
    "\n",
    "f7 = uproot.pandas.iterate(\"./data/data_other/gun_k0L_7gev_FTFP_BERT_5000evt_ILD_l5_v02steel.root\", fTreeName, branches)\n",
    "file7 = pd.concat(list(f7))\n",
    "\n",
    "f45 = uproot.pandas.iterate(\"./data/data_other/gun_k0L_45gev_FTFP_BERT_5000evt_ILD_l5_v02steel.root\", fTreeName, branches)\n",
    "file45 = pd.concat(list(f45))\n",
    "\n",
    "\n",
    "f120 = uproot.pandas.iterate(\"./data/data_other/gun_k0L_120gev_FTFP_BERT_5000evt_ILD_l5_v02steel.root\", fTreeName, branches)\n",
    "file120 = pd.concat(list(f120))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [],
   "source": [
    "f6 = uproot.pandas.iterate(\"./data/data_other/gun_k0L_6gev_FTFP_BERT_5000evt_ILD_l5_v02steel.root\", fTreeName, branches)\n",
    "file6 = pd.concat(list(f6))\n",
    "\n",
    "f8 = uproot.pandas.iterate(\"./data/data_other/gun_k0L_8gev_FTFP_BERT_5000evt_ILD_l5_v02steel.root\", fTreeName, branches)\n",
    "file8 = pd.concat(list(f8))\n",
    "\n",
    "f9 = uproot.pandas.iterate(\"./data/data_other/gun_k0L_9gev_FTFP_BERT_5000evt_ILD_l5_v02steel.root\", fTreeName, branches)\n",
    "file9 = pd.concat(list(f9))\n",
    "\n",
    "f12 = uproot.pandas.iterate(\"./data/data_other/gun_k0L_12gev_FTFP_BERT_5000evt_ILD_l5_v02steel.root\", fTreeName, branches)\n",
    "file12 = pd.concat(list(f12))\n",
    "\n",
    "f14 = uproot.pandas.iterate(\"./data/data_other/gun_k0L_14gev_FTFP_BERT_5000evt_ILD_l5_v02steel.root\", fTreeName, branches)\n",
    "file14 = pd.concat(list(f14))\n",
    "\n",
    "f16 = uproot.pandas.iterate(\"./data/data_other/gun_k0L_16gev_FTFP_BERT_5000evt_ILD_l5_v02steel.root\", fTreeName, branches)\n",
    "file16 = pd.concat(list(f16))\n",
    "\n",
    "f18 = uproot.pandas.iterate(\"./data/data_other/gun_k0L_18gev_FTFP_BERT_5000evt_ILD_l5_v02steel.root\", fTreeName, branches)\n",
    "file18 = pd.concat(list(f18))\n",
    "\n",
    "f35 = uproot.pandas.iterate(\"./data/data_other/gun_k0L_35gev_FTFP_BERT_5000evt_ILD_l5_v02steel.root\", fTreeName, branches)\n",
    "file35 = pd.concat(list(f35))\n",
    "\n",
    "f55 = uproot.pandas.iterate(\"./data/data_other/gun_k0L_55gev_FTFP_BERT_5000evt_ILD_l5_v02steel.root\", fTreeName, branches)\n",
    "file55 = pd.concat(list(f55))\n",
    "\n",
    "f65 = uproot.pandas.iterate(\"./data/data_other/gun_k0L_65gev_FTFP_BERT_5000evt_ILD_l5_v02steel.root\", fTreeName, branches)\n",
    "file65 = pd.concat(list(f65))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [],
   "source": [
    "out5 = xg_reg.predict( file5 )\n",
    "out10 = xg_reg.predict( file10 )\n",
    "out15 = xg_reg.predict( file15 )\n",
    "out20 = xg_reg.predict( file20 )\n",
    "out25 = xg_reg.predict( file25 )\n",
    "out30 = xg_reg.predict( file30 )\n",
    "out40 = xg_reg.predict( file40 )\n",
    "out50 = xg_reg.predict( file50 )\n",
    "out60 = xg_reg.predict( file60 )\n",
    "\n",
    "out_train = xg_reg.predict( x_train )\n",
    "out_test = xg_reg.predict( x_test )\n",
    "#out_test = xg_reg.predict( x_test[ branches ].values )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [],
   "source": [
    "out1 = xg_reg.predict( file1 )\n",
    "out2 = xg_reg.predict( file2 )\n",
    "out3 = xg_reg.predict( file3 )\n",
    "out4 = xg_reg.predict( file4 )\n",
    "out70 = xg_reg.predict( file70 )\n",
    "out80 = xg_reg.predict( file80 )\n",
    "out90 = xg_reg.predict( file90 )\n",
    "\n",
    "out7 = xg_reg.predict( file7 )\n",
    "out45 = xg_reg.predict( file45 )\n",
    "\n",
    "out120 = xg_reg.predict( file120 )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [],
   "source": [
    "out6 = xg_reg.predict( file6 )\n",
    "out8 = xg_reg.predict( file8 )\n",
    "out9 = xg_reg.predict( file9 )\n",
    "out12 = xg_reg.predict( file12 )\n",
    "out14 = xg_reg.predict( file14 )\n",
    "out16 = xg_reg.predict( file16 )\n",
    "out18 = xg_reg.predict( file18 )\n",
    "out35 = xg_reg.predict( file35 )\n",
    "out55 = xg_reg.predict( file55 )\n",
    "out65 = xg_reg.predict( file65 )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [],
   "source": [
    "bdt = ROOT.TFile('bdt.root', 'recreate')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [],
   "source": [
    "tree5 = ROOT.TTree( 'tree5', 'tree5' )\n",
    "tree10 = ROOT.TTree( 'tree10', 'tree10' )\n",
    "tree15 = ROOT.TTree( 'tree15', 'tree15' )\n",
    "tree20 = ROOT.TTree( 'tree20', 'tree20' )\n",
    "tree25 = ROOT.TTree( 'tree25', 'tree25' )\n",
    "tree30 = ROOT.TTree( 'tree30', 'tree30' )\n",
    "tree40 = ROOT.TTree( 'tree40', 'tree40' )\n",
    "tree50 = ROOT.TTree( 'tree50', 'tree50' )\n",
    "tree60 = ROOT.TTree( 'tree60', 'tree60' )\n",
    "\n",
    "tree_train = ROOT.TTree( 'tree_train', 'tree_train' )\n",
    "tree_test = ROOT.TTree( 'tree_test', 'tree_test' )\n",
    "\n",
    "before_train = ROOT.TTree( 'before_train', 'before_train' )\n",
    "before_test = ROOT.TTree( 'before_test', 'before_test' )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [],
   "source": [
    "tree1 = ROOT.TTree( 'tree1', 'tree1' )\n",
    "tree2 = ROOT.TTree( 'tree2', 'tree2' )\n",
    "tree3 = ROOT.TTree( 'tree3', 'tree3' )\n",
    "tree4 = ROOT.TTree( 'tree4', 'tree4' )\n",
    "tree70 = ROOT.TTree( 'tree70', 'tree70' )\n",
    "tree80 = ROOT.TTree( 'tree80', 'tree80' )\n",
    "tree90 = ROOT.TTree( 'tree90', 'tree90' )\n",
    "\n",
    "tree7 = ROOT.TTree( 'tree7', 'tree7' )\n",
    "tree45 = ROOT.TTree( 'tree45', 'tree45' )\n",
    "\n",
    "tree120 = ROOT.TTree( 'tree120', 'tree120' )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [],
   "source": [
    "tree6 = ROOT.TTree( 'tree6', 'tree6' )\n",
    "tree8 = ROOT.TTree( 'tree8', 'tree8' )\n",
    "tree9 = ROOT.TTree( 'tree9', 'tree9' )\n",
    "tree12 = ROOT.TTree( 'tree12', 'tree12' )\n",
    "tree14 = ROOT.TTree( 'tree14', 'tree14' )\n",
    "tree16 = ROOT.TTree( 'tree16', 'tree16' )\n",
    "tree18 = ROOT.TTree( 'tree18', 'tree18' )\n",
    "tree35 = ROOT.TTree( 'tree35', 'tree35' )\n",
    "tree55 = ROOT.TTree( 'tree55', 'tree55' )\n",
    "tree65 = ROOT.TTree( 'tree65', 'tree65' )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [],
   "source": [
    "bdt5 = array( 'f', [ 0.0 ] )\n",
    "bdt10 = array( 'f', [ 0.0 ] )\n",
    "bdt15 = array( 'f', [ 0.0 ] )\n",
    "bdt20 = array( 'f', [ 0.0 ] )\n",
    "bdt25 = array( 'f', [ 0.0 ] )\n",
    "bdt30 = array( 'f', [ 0.0 ] )\n",
    "bdt40 = array( 'f', [ 0.0 ] )\n",
    "bdt50 = array( 'f', [ 0.0 ] )\n",
    "bdt60 = array( 'f', [ 0.0 ] )\n",
    "\n",
    "bdt_train = array( 'f', [ 0.0 ] )\n",
    "bdt_test = array( 'f', [ 0.0 ] )\n",
    "\n",
    "train = array( 'f', [ 0.0 ] )\n",
    "test = array( 'f', [ 0.0 ] )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [],
   "source": [
    "bdt1 = array( 'f', [ 0.0 ] )\n",
    "bdt2 = array( 'f', [ 0.0 ] )\n",
    "bdt3 = array( 'f', [ 0.0 ] )\n",
    "bdt4 = array( 'f', [ 0.0 ] )\n",
    "bdt70 = array( 'f', [ 0.0 ] )\n",
    "bdt80 = array( 'f', [ 0.0 ] )\n",
    "bdt90 = array( 'f', [ 0.0 ] )\n",
    "\n",
    "bdt7 = array( 'f', [ 0.0 ] )\n",
    "bdt45 = array( 'f', [ 0.0 ] )\n",
    "\n",
    "bdt120 = array( 'f', [ 0.0 ] )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [],
   "source": [
    "bdt6 = array( 'f', [ 0.0 ] )\n",
    "bdt8 = array( 'f', [ 0.0 ] )\n",
    "bdt9 = array( 'f', [ 0.0 ] )\n",
    "bdt12 = array( 'f', [ 0.0 ] )\n",
    "bdt14 = array( 'f', [ 0.0 ] )\n",
    "bdt16 = array( 'f', [ 0.0 ] )\n",
    "bdt18 = array( 'f', [ 0.0 ] )\n",
    "bdt35 = array( 'f', [ 0.0 ] )\n",
    "bdt55 = array( 'f', [ 0.0 ] )\n",
    "bdt65 = array( 'f', [ 0.0 ] )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<cppyy.gbl.TBranch object at 0xdebbe50>"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tree5.Branch( 'bdt5', bdt5, 'bdt5/F' )\n",
    "tree10.Branch( 'bdt10', bdt10, 'bdt10/F' )\n",
    "tree15.Branch( 'bdt15', bdt15, 'bdt15/F' )\n",
    "tree20.Branch( 'bdt20', bdt20, 'bdt20/F' )\n",
    "tree25.Branch( 'bdt25', bdt25, 'bdt25/F' )\n",
    "tree30.Branch( 'bdt30', bdt30, 'bdt30/F' )\n",
    "tree40.Branch( 'bdt40', bdt40, 'bdt40/F' )\n",
    "tree50.Branch( 'bdt50', bdt50, 'bdt50/F' )\n",
    "tree60.Branch( 'bdt60', bdt60, 'bdt60/F' )\n",
    "\n",
    "tree_train.Branch( 'bdt_train', bdt_train, 'bdt_train/F' )\n",
    "tree_test.Branch( 'bdt_test', bdt_test, 'bdt_test/F' )\n",
    "\n",
    "before_train.Branch( 'train', train, 'train/F' )\n",
    "before_test.Branch( 'test', test, 'test/F' )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<cppyy.gbl.TBranch object at 0xdebfb10>"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tree1.Branch( 'bdt1', bdt1, 'bdt1/F' )\n",
    "tree2.Branch( 'bdt2', bdt2, 'bdt2/F' )\n",
    "tree3.Branch( 'bdt3', bdt3, 'bdt3/F' )\n",
    "tree4.Branch( 'bdt4', bdt4, 'bdt4/F' )\n",
    "tree70.Branch( 'bdt70', bdt70, 'bdt70/F' )\n",
    "tree80.Branch( 'bdt80', bdt80, 'bdt80/F' )\n",
    "tree90.Branch( 'bdt90', bdt90, 'bdt90/F' )\n",
    "\n",
    "tree7.Branch( 'bdt7', bdt7, 'bdt7/F' )\n",
    "tree45.Branch( 'bdt45', bdt45, 'bdt45/F' )\n",
    "\n",
    "tree120.Branch( 'bdt120', bdt120, 'bdt120/F' )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<cppyy.gbl.TBranch object at 0xdcf72f0>"
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tree6.Branch( 'bdt6', bdt6, 'bdt6/F' )\n",
    "tree8.Branch( 'bdt8', bdt8, 'bdt8/F' )\n",
    "tree9.Branch( 'bdt9', bdt9, 'bdt9/F' )\n",
    "tree12.Branch( 'bdt12', bdt12, 'bdt12/F' )\n",
    "tree14.Branch( 'bdt14', bdt14, 'bdt14/F' )\n",
    "tree16.Branch( 'bdt16', bdt16, 'bdt16/F' )\n",
    "tree18.Branch( 'bdt18', bdt18, 'bdt18/F' )\n",
    "tree35.Branch( 'bdt35', bdt35, 'bdt35/F' )\n",
    "tree55.Branch( 'bdt55', bdt55, 'bdt55/F' )\n",
    "tree65.Branch( 'bdt65', bdt65, 'bdt65/F' )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [],
   "source": [
    "#x_train_energy = x_train.energy.to_numpy(x_train.energy)\n",
    "#x_test_energy = x_test.energy.to_numpy(x_test.energy)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [],
   "source": [
    "for i in range( 0, len( out5 ) ):\n",
    "    bdt5[ 0 ] = out5[ i ]\n",
    "    tree5.Fill()\n",
    "\n",
    "for i in range( 0, len( out10 ) ):\n",
    "    bdt10[ 0 ] = out10[ i ]\n",
    "    tree10.Fill()\n",
    "\n",
    "for i in range( 0, len( out15 ) ):\n",
    "    bdt15[ 0 ] = out15[ i ]\n",
    "    tree15.Fill()\n",
    "\n",
    "for i in range( 0, len( out20 ) ):\n",
    "    bdt20[ 0 ] = out20[ i ]\n",
    "    tree20.Fill()\n",
    "\n",
    "for i in range( 0, len( out25 ) ):\n",
    "    bdt25[ 0 ] = out25[ i ]\n",
    "    tree25.Fill()\n",
    "\n",
    "for i in range( 0, len( out30 ) ):\n",
    "    bdt30[ 0 ] = out30[ i ]\n",
    "    tree30.Fill()\n",
    "\n",
    "for i in range( 0, len( out40 ) ):\n",
    "    bdt40[ 0 ] = out40[ i ]\n",
    "    tree40.Fill()\n",
    "\n",
    "for i in range( 0, len( out50 ) ):\n",
    "    bdt50[ 0 ] = out50[ i ]\n",
    "    tree50.Fill()\n",
    "\n",
    "for i in range( 0, len( out60 ) ):\n",
    "    bdt60[ 0 ] = out60[ i ]\n",
    "    tree60.Fill()\n",
    "    \n",
    "    \n",
    "for i in range( 0, len( out_train ) ):\n",
    "    bdt_train[ 0 ] = out_train[ i ]\n",
    "    tree_train.Fill()\n",
    "    \n",
    "for i in range( 0, len( out_test ) ):\n",
    "    bdt_test[ 0 ] = out_test[ i ]\n",
    "    tree_test.Fill()\n",
    "\n",
    "    \n",
    "#for i in range( 0, len( x_train_energy) ):\n",
    "#    train[ 0 ] = x_train_energy[ i ]\n",
    "#    before_train.Fill()\n",
    "    \n",
    "#for i in range( 0, len( x_test_energy ) ):\n",
    "#    test[ 0 ] = x_test_energy[ i ]\n",
    "#    before_test.Fill()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [],
   "source": [
    "for i in range( 0, len( out1 ) ):\n",
    "    bdt1[ 0 ] = out1[ i ]\n",
    "    tree1.Fill()\n",
    "\n",
    "for i in range( 0, len( out2 ) ):\n",
    "    bdt2[ 0 ] = out2[ i ]\n",
    "    tree2.Fill()\n",
    "\n",
    "for i in range( 0, len( out3 ) ):\n",
    "    bdt3[ 0 ] = out3[ i ]\n",
    "    tree3.Fill()\n",
    "\n",
    "for i in range( 0, len( out4 ) ):\n",
    "    bdt4[ 0 ] = out4[ i ]\n",
    "    tree4.Fill()\n",
    "\n",
    "for i in range( 0, len( out70 ) ):\n",
    "    bdt70[ 0 ] = out70[ i ]\n",
    "    tree70.Fill()\n",
    "\n",
    "for i in range( 0, len( out80 ) ):\n",
    "    bdt80[ 0 ] = out80[ i ]\n",
    "    tree80.Fill()\n",
    "\n",
    "for i in range( 0, len( out90 ) ):\n",
    "    bdt90[ 0 ] = out90[ i ]\n",
    "    tree90.Fill()\n",
    "\n",
    "    \n",
    "for i in range( 0, len( out7 ) ):\n",
    "    bdt7[ 0 ] = out7[ i ]\n",
    "    tree7.Fill()\n",
    "\n",
    "for i in range( 0, len( out45 ) ):\n",
    "    bdt45[ 0 ] = out45[ i ]\n",
    "    tree45.Fill()\n",
    "    \n",
    "\n",
    "for i in range( 0, len( out120 ) ):\n",
    "    bdt120[ 0 ] = out120[ i ]\n",
    "    tree120.Fill()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [],
   "source": [
    "for i in range( 0, len( out6 ) ):\n",
    "    bdt6[ 0 ] = out6[ i ]\n",
    "    tree6.Fill()\n",
    "    \n",
    "for i in range( 0, len( out8 ) ):\n",
    "    bdt8[ 0 ] = out8[ i ]\n",
    "    tree8.Fill()\n",
    "    \n",
    "for i in range( 0, len( out9 ) ):\n",
    "    bdt9[ 0 ] = out9[ i ]\n",
    "    tree9.Fill()\n",
    "    \n",
    "for i in range( 0, len( out12 ) ):\n",
    "    bdt12[ 0 ] = out12[ i ]\n",
    "    tree12.Fill()\n",
    "    \n",
    "for i in range( 0, len( out14 ) ):\n",
    "    bdt14[ 0 ] = out14[ i ]\n",
    "    tree14.Fill()\n",
    "    \n",
    "for i in range( 0, len( out16 ) ):\n",
    "    bdt16[ 0 ] = out16[ i ]\n",
    "    tree16.Fill()\n",
    "    \n",
    "for i in range( 0, len( out18 ) ):\n",
    "    bdt18[ 0 ] = out18[ i ]\n",
    "    tree18.Fill()\n",
    "    \n",
    "for i in range( 0, len( out35 ) ):\n",
    "    bdt35[ 0 ] = out35[ i ]\n",
    "    tree35.Fill()\n",
    "    \n",
    "for i in range( 0, len( out55 ) ):\n",
    "    bdt55[ 0 ] = out55[ i ]\n",
    "    tree55.Fill()\n",
    "    \n",
    "for i in range( 0, len( out65 ) ):\n",
    "    bdt65[ 0 ] = out65[ i ]\n",
    "    tree65.Fill()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [],
   "source": [
    "bdt.Write()\n",
    "bdt.Close()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}