visualization.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Protein-Protein Binding Affinity Prediction using Random Forest Regressor\n",
    "\n",
    "##### Beryl Ramadhian Aribowo \n",
    "##### DDP - Master, Kanazawa University\n",
    "\n",
    "https://github.com/berylgithub"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Feature Extraction\n",
    "\n",
    "The data is pre-processed to acquire the $D=\\{(y^{(n)}, \\vec{x}^{(n)})\\}^N_{n=1}$ array. By calculating the Heaviside step function that counts the contacts of atoms between proteins ($d_{cutoff}$) within 12 Angstrom. The formula to count the number of occurence between $j$ and $i$ atom is as follows:\n",
    "<br>\n",
    "$x_{Z(P_X(j)),Z(P_Y(i))}\\equiv\\sum^{K_j}_{k=1}\\sum^{L_i}_{l=1} \\odot (d_{cutoff} - d_{kl})$\n",
    "<br>\n",
    "\n",
    "Here $P_X(j)$ corresponds to the protein atoms within the first chain of the protein pairing combination, and $P_Y(i)$ corresponds to the protein atoms within the second one, where four atom types are considered respectively as follows:\n",
    "<br>\n",
    "$\\{P_N(j)\\}_{j=1}^{4} = \\{C,N,O,S\\}$, $N \\in Z^+$,<br>\n",
    "The hydrophobic patches and acid patches interactions respectively are also considered: \n",
    "<br>\n",
    "- hydrophobic patch : $\\{C\\alpha_H\\}$\n",
    "<br>\n",
    "- acid patch : $\\{C\\alpha_A\\}$,\n",
    "<br>\n",
    "\n",
    "where $C\\alpha_H$ and $C\\alpha_A$ is the total interactions between carbon-alpha on Hydrophobic and Acid patch respectively.\n",
    "\n",
    "Therefore the total atomic interaction features used are $|\\{P_N(j)\\}_{j=1}^{4}|^2 + |\\{C\\alpha_H\\}| + |\\{C\\alpha_A\\}| = 18$\n",
    "\n",
    "Suppose there are four chains of protein $S = \\{P_1, P_2, P_3, P_4 \\}$, the interaction is the combinations of two chains, hence the total of interactions between two chains for this case is $c^{|S|}_2$=6 (where $c$ here refers to combination formula), the list of chains combination are as follows:\n",
    "<br>\n",
    "$\\sum interaction = interaction(P_1, P_2) + interaction(P_1,P_3) + interaction(P_1,P_4) + interaction(P_2,P_3) + interaction(P_2,P_4) + interaction(P_3,P_4)$\n",
    "\n",
    "The script used for data preprocessing and feature extraction is available at https://github.com/berylgithub/ppbap."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Data and Model loading"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "from sklearn.ensemble import RandomForestRegressor\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "from sklearn.model_selection import KFold\n",
    "from sklearn.metrics import mean_squared_error\n",
    "from sklearn.inspection import permutation_importance\n",
    "from scipy.stats import pearsonr\n",
    "from math import sqrt\n",
    "\n",
    "import pickle\n",
    "import os"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### First type model: 16 features ([C, N, O, S] interactions)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training Features Shape: (1812, 16)\n",
      "Training Labels Shape: (1812,)\n",
      "Testing Features Shape: (604, 16)\n",
      "Testing Labels Shape: (604,)\n"
     ]
    }
   ],
   "source": [
    "'''Dataset Loader'''\n",
    "def dataset_loader(filepath):\n",
    "    data = []\n",
    "    try:\n",
    "        with open(filepath, 'rb') as fr:\n",
    "            try:\n",
    "                while True:\n",
    "                    data.append(pickle.load(fr))\n",
    "            except EOFError:\n",
    "                pass            \n",
    "    except FileNotFoundError:\n",
    "        print('File is not found')\n",
    "    saved_ids = [d['id'] for d in data]\n",
    "    return data\n",
    "\n",
    "standard_data_file = os.getcwd()+'/Data/dataset_alpha_12A.pkl'\n",
    "combined_ha_data_file = os.getcwd()+'/Data/dataset_ha_beta_12A.pkl'\n",
    "dataset = dataset_loader(standard_data_file)\n",
    "\n",
    "features = np.array([data['x_vector'] for data in dataset])\n",
    "labels = np.array([data['y'] for data in dataset])\n",
    "x_train, x_test, y_train, y_test = train_test_split(features, labels, test_size=0.25, random_state=13)\n",
    "print('Training Features Shape:', x_train.shape)\n",
    "print('Training Labels Shape:', y_train.shape)\n",
    "print('Testing Features Shape:', x_test.shape)\n",
    "print('Testing Labels Shape:', y_test.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\beryl\\Anaconda3\\lib\\site-packages\\sklearn\\utils\\deprecation.py:144: FutureWarning: The sklearn.ensemble.forest module is  deprecated in version 0.22 and will be removed in version 0.24. The corresponding classes / functions should instead be imported from sklearn.ensemble. Anything that cannot be imported from sklearn.ensemble is now part of the private API.\n",
      "  warnings.warn(message, FutureWarning)\n",
      "C:\\Users\\beryl\\Anaconda3\\lib\\site-packages\\sklearn\\utils\\deprecation.py:144: FutureWarning: The sklearn.tree.tree module is  deprecated in version 0.22 and will be removed in version 0.24. The corresponding classes / functions should instead be imported from sklearn.tree. Anything that cannot be imported from sklearn.tree is now part of the private API.\n",
      "  warnings.warn(message, FutureWarning)\n",
      "C:\\Users\\beryl\\Anaconda3\\lib\\site-packages\\sklearn\\base.py:318: UserWarning: Trying to unpickle estimator DecisionTreeRegressor from version 0.21.2 when using version 0.22.1. This might lead to breaking code or invalid results. Use at your own risk.\n",
      "  UserWarning)\n",
      "C:\\Users\\beryl\\Anaconda3\\lib\\site-packages\\sklearn\\base.py:318: UserWarning: Trying to unpickle estimator RandomForestRegressor from version 0.21.2 when using version 0.22.1. This might lead to breaking code or invalid results. Use at your own risk.\n",
      "  UserWarning)\n"
     ]
    }
   ],
   "source": [
    "'''Model Loader'''\n",
    "standard_model_file= os.getcwd()+\"/Model/rf_pp_alpha_12A_split.pkl\"\n",
    "combined_ha_model_file= os.getcwd()+\"/Model/rf_pp_ha_beta_12A_split.pkl\"\n",
    "with open(standard_model_file, \"rb\") as f:\n",
    "    rf = pickle.load(f)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=12)]: Using backend ThreadingBackend with 12 concurrent workers.\n",
      "[Parallel(n_jobs=12)]: Done  26 tasks      | elapsed:    0.0s\n",
      "[Parallel(n_jobs=12)]: Done 176 tasks      | elapsed:    0.0s\n",
      "[Parallel(n_jobs=12)]: Done 426 tasks      | elapsed:    0.0s\n",
      "[Parallel(n_jobs=12)]: Done 776 tasks      | elapsed:    0.0s\n",
      "[Parallel(n_jobs=12)]: Done 1000 out of 1000 | elapsed:    0.0s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean Absolute Error: 0.52\n",
      "Accuracy: 91.89 %.\n",
      "Root Mean Squared Error : 0.66\n",
      "Pearson Correlation Coefficient : 0.98\n"
     ]
    }
   ],
   "source": [
    "'''Model evaluation on training set'''\n",
    "#Mean Absolute Error\n",
    "preds = rf.predict(x_train)\n",
    "errors = abs(preds - y_train)\n",
    "print('Mean Absolute Error:', round(np.mean(errors), 2))\n",
    "\n",
    "#Mean Absolute Percentage Error & Accuracy\n",
    "mape = 100 * (errors / y_train)\n",
    "accuracy = 100 - np.mean(mape)\n",
    "print('Accuracy:', round(accuracy, 2), '%.')\n",
    "\n",
    "#Root Mean Squared Error\n",
    "rmse = sqrt(mean_squared_error(y_train, preds))\n",
    "print('Root Mean Squared Error :', round(rmse, 2))\n",
    "\n",
    "#Pearson Correlation Coefficient (PCC) score\n",
    "pcc = pearsonr(y_train, preds)\n",
    "print('Pearson Correlation Coefficient :', round(pcc[0],2))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "'''Visualization on training set'''\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "fig_size = np.zeros(2)\n",
    "fig_size[0] = 15\n",
    "fig_size[1] = 10\n",
    "plt.rcParams[\"figure.figsize\"] = fig_size\n",
    "\n",
    "f, ax = plt.subplots()\n",
    "\n",
    "plt.plot(y_train, preds, 'o', markerfacecolor=\"None\", markeredgecolor='green', markeredgewidth=1)\n",
    "#ax.plot(ax.get_xlim(), ax.get_ylim(), ls=\"--\", c=\".3\")\n",
    "coef = np.polyfit(y_train,preds,1)\n",
    "poly1d_fn = np.poly1d(coef)\n",
    "plt.plot(y_train, poly1d_fn(y_train), '--k')\n",
    "\n",
    "plt.title(\"Training data binding affinity\")\n",
    "plt.xlabel('Actual binding affinity data')\n",
    "plt.ylabel('Predicted binding affinity data')\n",
    "#plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=12)]: Using backend ThreadingBackend with 12 concurrent workers.\n",
      "[Parallel(n_jobs=12)]: Done  26 tasks      | elapsed:    0.0s\n",
      "[Parallel(n_jobs=12)]: Done 176 tasks      | elapsed:    0.0s\n",
      "[Parallel(n_jobs=12)]: Done 426 tasks      | elapsed:    0.0s\n",
      "[Parallel(n_jobs=12)]: Done 776 tasks      | elapsed:    0.0s\n",
      "[Parallel(n_jobs=12)]: Done 1000 out of 1000 | elapsed:    0.0s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean Absolute Error: 1.47\n",
      "Accuracy: 76.67 %.\n",
      "Root Mean Squared Error : 1.87\n",
      "Pearson Correlation Coefficient : 0.35\n"
     ]
    }
   ],
   "source": [
    "'''Model evaluation on Test set'''\n",
    "#Mean Absolute Error\n",
    "preds = rf.predict(x_test)\n",
    "errors = abs(preds - y_test)\n",
    "print('Mean Absolute Error:', round(np.mean(errors), 2))\n",
    "\n",
    "#Mean Absolute Percentage Error & Accuracy\n",
    "mape = 100 * (errors / y_test)\n",
    "accuracy = 100 - np.mean(mape)\n",
    "print('Accuracy:', round(accuracy, 2), '%.')\n",
    "\n",
    "#Root Mean Squared Error\n",
    "rmse = sqrt(mean_squared_error(y_test, preds))\n",
    "print('Root Mean Squared Error :', round(rmse, 2))\n",
    "\n",
    "#Pearson Correlation Coefficient (PCC) score\n",
    "pcc = pearsonr(y_test, preds)\n",
    "print('Pearson Correlation Coefficient :', round(pcc[0],2))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "'''Visualization on test set'''\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "fig_size = np.zeros(2)\n",
    "fig_size[0] = 15\n",
    "fig_size[1] = 10\n",
    "plt.rcParams[\"figure.figsize\"] = fig_size\n",
    "\n",
    "f, ax = plt.subplots()\n",
    "\n",
    "plt.plot(y_test, preds, 'o', markerfacecolor=\"None\", markeredgecolor='green', markeredgewidth=1)\n",
    "#ax.plot(ax.get_xlim(), ax.get_ylim(), ls=\"--\", c=\".3\")\n",
    "coef = np.polyfit(y_test,preds,1)\n",
    "poly1d_fn = np.poly1d(coef)\n",
    "plt.plot(y_test, poly1d_fn(y_test), '--k')\n",
    "\n",
    "\n",
    "plt.title(\"Test data binding affinity\")\n",
    "plt.xlabel('Actual binding affinity data')\n",
    "plt.ylabel('Predicted binding affinity data')\n",
    "#plt.legend()\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Top Feature ranking:\n",
      "1. feature 10 (0.083682)\n",
      "2. feature 13 (0.077048)\n",
      "3. feature 0 (0.073376)\n",
      "4. feature 8 (0.069573)\n",
      "5. feature 12 (0.066819)\n",
      "6. feature 9 (0.063811)\n",
      "7. feature 7 (0.063336)\n",
      "8. feature 3 (0.063243)\n",
      "9. feature 2 (0.060709)\n",
      "10. feature 14 (0.059633)\n",
      "11. feature 6 (0.058835)\n",
      "12. feature 4 (0.058746)\n",
      "13. feature 5 (0.054110)\n",
      "14. feature 11 (0.050171)\n",
      "15. feature 1 (0.049238)\n",
      "16. feature 15 (0.047671)\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA34AAAJcCAYAAACmOnadAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjAsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+17YcXAAAgAElEQVR4nOzde5xlZ1kn+t9DGgIChmPSXBNJIAHtICIUgXMOKseMShQIOvGQoICaY2AkOqiokdERM+OMeHQyesALGDTGSwJxxOYA4oxRHFFDKhguAQKdgKYNSkPCJSCEhGf+2KuxqFSldye1q6rffL+fz/70Xmu9a6/nrb1X7fr1uy7V3QEAAGBcd9vqAgAAAFgswQ8AAGBwgh8AAMDgBD8AAIDBCX4AAACDE/wAAAAGJ/gBwDqq6teq6qe2ug4AuLPKffwA2GhV9cEkD0hy64rZj+ju6+/Eaz45ye9099F3rrpDU1X9VpK93f2TW10LAIceI34ALMrTuvs+Kx53OPRthKrasZXbvzOq6rCtrgGAQ5vgB8CmqqonVtVfVdXHqurt00je/mXfU1XvqapPVtW1VfW8af69k7wxyYOr6qbp8eCq+q2q+o8r1n9yVe1dMf3BqvrxqnpHkk9V1Y5pvT+oqn1V9YGq+sHbqfULr7//tavqx6rqw1X1oap6RlV9S1W9r6puqKoXr1j3JVV1SVVdPPXnbVX11SuWf2VV/fn0c7iqqp6+aru/WlVvqKpPJTkzyXcm+bGp76+b2p1TVddMr//uqvq2Fa/x3VX1l1X1C1V149TXU1Ys/7Kq+s2qun5a/toVy55aVVdOtf1VVT16xbIfr6p/mLZ5dVWdPMfbDsAWE/wA2DRV9ZAkr0/yH5N8WZIXJfmDqto5Nflwkqcm+dIk35PkvKp6bHd/KskpSa6/AyOIZyT51iT3S/L5JK9L8vYkD0lycpIXVtU3z/laD0xyz2ndf5/klUm+K8njknxtkn9fVQ9b0f7UJK+Z+vp7SV5bVXevqrtPdfxJkvsn+YEkv1tVj1yx7rOS/GyS+yb57SS/m+Tnp74/bWpzzbTdI5L8TJLfqaoHrXiNJyS5OslRSX4+yflVVdOyC5N8SZITpxrOS5KqemySVyV5XpIjk/x6kt1VdfhU39lJHt/d903yzUk+OOfPDoAtJPgBsCivnUaMPrZiNOm7kryhu9/Q3Z/v7v+eZDnJtyRJd7++u6/pmTdnFoy+9k7W8cvdfV13/3OSxyfZ2d3ndvfN3X1tZuHt9Dlf63NJfra7P5fkoswC1S919ye7+6okVyV59Ir2V3T3JVP7/5JZaHzi9LhPkp+b6rg0yf+fWUjd74+6+y3Tz+kzaxXT3a/p7uunNhcneX+Sk1Y0+bvufmV335rkgiQPSvKAKRyekuT53X1jd39u+nknyfcl+fXuvqy7b+3uC5J8dqr51iSHJ9lVVXfv7g929zVz/uwA2EKCHwCL8ozuvt/0eMY076FJvmNFIPxYkidlFkhSVadU1d9Mh01+LLNAeNSdrOO6Fc8fmtnhoiu3/+LMLkQzj49OISpJ/nn6959WLP/nzALdbbbd3Z9PsjfJg6fHddO8/f4us5HEtepeU1U9Z8UhmR9L8qh88c/rH1ds/9PT0/skOSbJDd194xov+9AkP7LqZ3RMkgd3954kL0zykiQfrqqLqurBB6oTgK0n+AGwma5LcuGKQHi/7r53d/9cVR2e5A+S/EKSB3T3/ZK8Icn+QxPXugz1pzI7XHG/B67RZuV61yX5wKrt37e7v+VO92xtx+x/UlV3S3J0kuunxzHTvP2+PMk/rFP3baar6qGZjVaeneTI6ef1rvzLz+v2XJfky6rqfuss+9lVP6Mv6e7fT5Lu/r3uflJmAbGTvHSO7QGwxQQ/ADbT7yR5WlV9c1UdVlX3nC6acnSSe2R2GOG+JLdMFyL5phXr/lOSI6vqiBXzrkzyLdOFSh6Y2WjU7Xlrkk9MFyi511TDo6rq8RvWwy/2uKr69ppdUfSFmR0y+TdJLssstP7YdM7fk5M8LbPDR9fzT0lWnj9478yC175kdmGczEb8Dqi7P5TZxXJ+par+t6mGr5sWvzLJ86vqCTVz76r61qq6b1U9sqq+YQrpn8lshPPWdTYDwDYi+AGwabr7uswuePLizALLdUl+NMnduvuTSX4wyauT3JjZxU12r1j3vUl+P8m10yGID87sAiVvz+wCI3+S5OIDbP/WzALWY5J8IMlHkvxGZhdHWYQ/SvLMzPrz7CTfPp1Pd3OSp2d2nt1HkvxKkudMfVzP+ZmdW/exqnptd787yS8m+evMQuFXJXnLQdT27MzOWXxvZhfVeWGSdPdyZuf5vWyqe0+S757WOTzJz001/2NmF4V5cQDY9tzAHQAWoKpekuT47v6ura4FAIz4AQAADE7wAwAAGJxDPQEAAAZnxA8AAGBwO7a6gI1y1FFH9bHHHrvVZQAAAGyJK6644iPdvXOtZcMEv2OPPTbLy8tbXQYAAMCWqKq/W2+ZQz0BAAAGJ/gBAAAMTvADAAAYnOAHAAAwOMEPAABgcIIfAADA4AQ/AACAwQl+AAAAgxP8AAAABif4AQAADE7wAwAAGJzgBwAAMDjBDwAAYHCCHwAAwOAEPwAAgMEJfgAAAIMT/AAAAAYn+AEAAAxO8AMAABic4AcAADA4wQ8AAGBwgh8AAMDgBD8AAIDBCX4AAACD27HVBRzyqra6gvV1b3UFAADANmDEDwAAYHCCHwAAwOAEPwAAgMEJfgAAAIMT/AAAAAYn+AEAAAxO8AMAABic4AcAADA4wQ8AAGBwgh8AAMDgBD8AAIDBCX4AAACDE/wAAAAGJ/gBAAAMTvADAAAYnOAHAAAwOMEPAABgcIIfAADA4AQ/AACAwS00+FXVU6rq6qraU1XnrLH88Kq6eFp+WVUdO82/e1VdUFXvrKr3VNVPLLJOAACAkS0s+FXVYUlenuSUJLuSnFFVu1Y1OzPJjd19fJLzkrx0mv8dSQ7v7q9K8rgkz9sfCgEAADg4ixzxOynJnu6+trtvTnJRklNXtTk1yQXT80uSnFxVlaST3LuqdiS5V5Kbk3xigbUCAAAMa5HB7yFJrlsxvXeat2ab7r4lyceTHJlZCPxUkg8l+fskv9DdN6zeQFWdVVXLVbW8b9++je8BAADAABYZ/GqNeT1nm5OS3JrkwUmOS/IjVfWw2zTsfkV3L3X30s6dO+9svQAAAENaZPDbm+SYFdNHJ7l+vTbTYZ1HJLkhybOS/HF3f667P5zkLUmWFlgrAADAsBYZ/C5PckJVHVdV90hyepLdq9rsTvLc6flpSS7t7s7s8M5vqJl7J3likvcusFYAAIBhLSz4TefsnZ3kTUnek+TV3X1VVZ1bVU+fmp2f5Miq2pPkh5Psv+XDy5PcJ8m7MguQv9nd71hUrQAAACOr2QDboW9paamXl5c3f8O11mmK28Qg7y0AAHBgVXVFd695itxCb+AOAADA1hP8AAAABif4AQAADE7wAwAAGJzgBwAAMDjBDwAAYHCCHwAAwOAEPwAAgMEJfgAAAIMT/AAAAAYn+AEAAAxO8AMAABjcjq0ugG2gaqsrWF/3VlcAAACHPCN+AAAAgxP8AAAABif4AQAADE7wAwAAGJzgBwAAMDjBDwAAYHCCHwAAwOAEPwAAgMEJfgAAAIMT/AAAAAYn+AEAAAxO8AMAABic4AcAADA4wQ8AAGBwgh8AAMDgBD8AAIDBCX4AAACDE/wAAAAGJ/gBAAAMTvADAAAYnOAHAAAwOMEPAABgcDu2ugDYEFVbXcHt697qCgAAuAsz4gcAADA4wQ8AAGBwgh8AAMDgBD8AAIDBCX4AAACDE/wAAAAGJ/gBAAAMTvADAAAYnOAHAAAwOMEPAABgcIIfAADA4AQ/AACAwQl+AAAAgxP8AAAABif4AQAADE7wAwAAGJzgBwAAMLiFBr+qekpVXV1Ve6rqnDWWH15VF0/LL6uqY6f531lVV654fL6qHrPIWgEAAEa1sOBXVYcleXmSU5LsSnJGVe1a1ezMJDd29/FJzkvy0iTp7t/t7sd092OSPDvJB7v7ykXVCgAAMLJFjvidlGRPd1/b3TcnuSjJqavanJrkgun5JUlOrqpa1eaMJL+/wDoBAACGtsjg95Ak162Y3jvNW7NNd9+S5ONJjlzV5plZJ/hV1VlVtVxVy/v27duQogEAAEazyOC3euQuSfpg2lTVE5J8urvftdYGuvsV3b3U3Us7d+6845UCAAAMbMcCX3tvkmNWTB+d5Pp12uytqh1Jjkhyw4rlp8dhntyV3OZI522kV/+/DQAAh4pFjvhdnuSEqjququ6RWYjbvarN7iTPnZ6fluTS7tlfl1V1tyTfkdm5gQAAANxBCxvx6+5bqursJG9KcliSV3X3VVV1bpLl7t6d5PwkF1bVnsxG+k5f8RJfl2Rvd1+7qBoBAADuCqoHOXxraWmpl5eXN3/DIxyapw+LN0I/BvldAQAwqqq6oruX1lq20Bu4AwAAsPUEPwAAgMEJfgAAAIMT/AAAAAYn+AEAAAxO8AMAABic4AcAADA4wQ8AAGBwgh8AAMDgBD8AAIDBCX4AAACDE/wAAAAGJ/gBAAAMbsdWFwAMpmqrK1hf91ZXAACwJYz4AQAADE7wAwAAGJzgBwAAMDjBDwAAYHCCHwAAwOAEPwAAgMEJfgAAAIMT/AAAAAYn+AEAAAxO8AMAABic4AcAADA4wQ8AAGBwgh8AAMDgBD8AAIDBCX4AAACDE/wAAAAGJ/gBAAAMTvADAAAY3I6tLgBg26na6grW173VFQAAhyAjfgAAAIMT/AAAAAbnUE+AEW3nw1UTh6wCwCYT/ADYvrZzgBVeATiEONQTAABgcIIfAADA4AQ/AACAwQl+AAAAgxP8AAAABif4AQAADE7wAwAAGJzgBwAAMDjBDwAAYHCCHwAAwOAEPwAAgMEJfgAAAIMT/AAAAAYn+AEAAAxux1YXAABDq9rqCtbXvdUVALBJjPgBAAAMbqHBr6qeUlVXV9WeqjpnjeWHV9XF0/LLqurYFcseXVV/XVVXVdU7q+qei6wVAABgVAsLflV1WJKXJzklya4kZ1TVrlXNzkxyY3cfn+S8JC+d1t2R5HeSPL+7T0zy5CSfW1StAAAAI1vkiN9JSfZ097XdfXOSi5KcuqrNqUkumJ5fkuTkqqok35TkHd399iTp7o92960LrBUAAGBYiwx+D0ly3YrpvdO8Ndt09y1JPp7kyCSPSNJV9aaqeltV/dhaG6iqs6pquaqW9+3bt+EdAAAAGMEig99alzFbffmw9drsSPKkJN85/fttVXXybRp2v6K7l7p7aefOnXe2XgAAgCEtMvjtTXLMiumjk1y/XpvpvL4jktwwzX9zd3+kuz+d5A1JHrvAWgGA9VRt3wcAc1lk8Ls8yQlVdVxV3SPJ6Ul2r2qzO8lzp+enJbm0uzvJm5I8uqq+ZAqEX5/k3QusFQAAYFgLu4F7d99SVWdnFuIOS/Kq7r6qqs5Nstzdu5Ocn+TCqtqT2Ujf6dO6N1bVf8ksPHaSN3T36xdVKwAAwMhqNsB26FtaWurl5eXN3/B2Psxk3vdWHxZvhH7ow/YwQh+SMfqhD9vDIH/HAGyEqrqiu5fWWrbQG7gDAACw9RZ2qCcAwLaxnUctk/lGLkfoA7BljPgBAAAMTvADAAAYnOAHAAAwOMEPAABgcIIfAADA4AQ/AACAwQl+AAAAgxP8AAAABif4AQAADE7wAwAAGJzgBwAAMDjBDwAAYHCCHwAAwOAEPwAAgMEJfgAAAIMT/AAAAAYn+AEAAAxO8AMAABic4AcAADA4wQ8AAGBwgh8AAMDgBD8AAIDBCX4AAACDE/wAAAAGJ/gBAAAMTvADAAAYnOAHAAAwOMEPAABgcDu2ugAAAO5Cqra6gvV1b3UFsDBG/AAAAAYn+AEAAAxO8AMAABic4AcAADA4wQ8AAGBwgh8AAMDgBD8AAIDBCX4AAACDE/wAAAAGJ/gBAAAMTvADAAAYnOAHAAAwOMEPAABgcIIfAADA4AQ/AACAwQl+AAAAgxP8AAAABif4AQAADE7wAwAAGJzgBwAAMDjBDwAAYHAHDH5V9YCqOr+q3jhN76qqM+d58ap6SlVdXVV7quqcNZYfXlUXT8svq6pjp/nHVtU/V9WV0+PXDq5bAAAA7DfPiN9vJXlTkgdP0+9L8sIDrVRVhyV5eZJTkuxKckZV7VrV7MwkN3b38UnOS/LSFcuu6e7HTI/nz1EnAAAAa5gn+B3V3a9O8vkk6e5bktw6x3onJdnT3dd2981JLkpy6qo2pya5YHp+SZKTq6rmqhwAALZC1fZ9wDrmCX6fqqojk3SSVNUTk3x8jvUekuS6FdN7p3lrtpkC5ceTHDktO66q/raq3lxVX7vWBqrqrKparqrlffv2zVESAADAXc+OOdr8cJLdSR5eVW9JsjPJaXOst9Z/OfScbT6U5Mu7+6NV9bgkr62qE7v7E1/UsPsVSV6RJEtLS6tfGwAAgMwR/Lr7bVX19UkemVlQu7q7PzfHa+9NcsyK6aOTXL9Om71VtSPJEUlu6O5O8tlp+1dU1TVJHpFkeY7tAgAAsMIBg19VPWfVrMdWVbr7tw+w6uVJTqiq45L8Q5LTkzxrVZvdSZ6b5K8zG0W8tLu7qnZmFgBvraqHJTkhybUH7g4AAACrzXOo5+NXPL9nkpOTvC3J7Qa/7r6lqs7O7IqghyV5VXdfVVXnJlnu7t1Jzk9yYVXtSXJDZuEwSb4uyblVtf9CMs/v7hsOol8AAABManZU5UGsUHVEkgu7++mLKemOWVpa6uXlLTgSdDtfPWne91YfFm+EfujD9jBCH5Ix+qEP28MIfUjm68cIfUi2dz/0Yfs4yIzCTFVd0d1Lay2b56qeq306s0MvAQAAOATMc47f6/IvV+O8W2Y3Y3/1IosCAABg48xzjt8vrHh+S5K/6+69C6oHAACADTbP7RzevBmFAAAAsBjrBr+q+mRue8P1ZHYvv+7uL11YVQAAAGyYdYNfd993MwsBAABgMeY5xy9JUlX3z+w+fkmS7v77hVQEAADAhjrg7Ryq6ulV9f4kH0jy5iQfTPLGBdcFAADABpnnPn7/IckTk7yvu49LcnKStyy0KgAAADbMPMHvc9390SR3q6q7dfefJXnMgusCAABYX9X2fmwz85zj97Gquk+Sv0jyu1X14czu5wcAAMAhYJ4Rv1OTfDrJDyX54yTXJHnaIosCAABg48wz4ndWktd0994kFyy4HgAAADbYPCN+X5rkTVX1P6vqBVX1gEUXBQAAwMY5YPDr7p/p7hOTvCDJg5O8uar+x8IrAwAAYEPMM+K334eT/GOSjya5/2LKAQAAYKPNcwP3f1NVf57kT5McleT7uvvRiy4MAACAjTHPxV0emuSF3X3loosBAABg4x0w+HX3OZtRCAAAAItxMOf4AQAAcAgS/AAAAAY3V/CrqodW1b+ant+rqu672LIAAADYKPNc1fP7klyS5NenWUcnee0iiwIAAGDjzDPi94Ik/2eSTyRJd78/7uMHAABwyJgn+H22u2/eP1FVO5L04koCAABgI80T/N5cVS9Ocq+q+sYkr0nyusWWBQAAwEaZJ/idk2RfkncmeV6SNyT5yUUWBQAAwMY54A3ck9wryau6+5VJUlWHTfM+vcjCAAAA2BjzjPj9aWZBb797JfkfiykHAACAjTZP8Ltnd9+0f2J6/iWLKwkAAICNNE/w+1RVPXb/RFU9Lsk/L64kAAAANtI85/i9MMlrqur6afpBSZ65uJIAAADYSAcMft19eVV9RZJHJqkk7+3uzy28MgAAADbEPCN+SfL4JMdO7b+mqtLdv72wqgAAANgwBwx+VXVhkocnuTLJrdPsTiL4AQAAHALmGfFbSrKru3vRxQAAALDx5rmq57uSPHDRhQAAALAY84z4HZXk3VX11iSf3T+zu5++sKoAAADYMPMEv5csuggAAAAWZ57bObx5MwoBAABgMQ54jl9VPbGqLq+qm6rq5qq6tao+sRnFAQAAcOfNc3GXlyU5I8n7k9wryf8zzQMAAOAQMNcN3Lt7T1Ud1t23JvnNqvqrBdcFAADABpkn+H26qu6R5Mqq+vkkH0py78WWBQAAwEaZ51DPZ0/tzk7yqSTHJPn2RRYFAADAxpkn+D2juz/T3Z/o7p/p7h9O8tRFFwYAAMDGmCf4PXeNed+9wXUAAACwIOue41dVZyR5VpKHVdXuFYvum+Sjiy4MAACAjXF7F3f5q8wu5HJUkl9cMf+TSd6xyKIAAADYOOsGv+7+u6ram+RT3f3mTawJAACADXS75/hN9+37dFUdsUn1AAAAsMHmuY/fZ5K8s6r+e2a3c0iSdPcPLqwqAAAANsw8we/10wMAAIBD0AFv59DdFyT5/SRXTI/fm+YdUFU9paqurqo9VXXOGssPr6qLp+WXVdWxq5Z/eVXdVFUvmmd7AAAA3NYBg19VPTnJ+5O8PMmvJHlfVX3dHOsdNq1zSpJdSc6oql2rmp2Z5MbuPj7JeUleumr5eUneeKBtAQAAsL55DvX8xSTf1N1XJ0lVPSKzEcDHHWC9k5Ls6e5rp/UuSnJqknevaHNqkpdMzy9J8rKqqu7uqnpGkmuz4rxCAAAADt4BR/yS3H1/6EuS7n5fkrvPsd5Dkly3YnrvNG/NNt19S5KPJzmyqu6d5MeT/MztbaCqzqqq5apa3rdv3xwlAQAA3PXME/yWq+r8qnry9HhlZuf6HUitMa/nbPMzSc7r7ptubwPd/YruXurupZ07d85REgAAwF3PPId6/pskL0jyg5kFtb/I7Fy/A9mb5JgV00cnuX6dNnurakeSI5LckOQJSU6rqp9Pcr8kn6+qz3T3y+bYLgAAACscMPh192er6mVJ/jTJ55Nc3d03z/Halyc5oaqOS/IPSU5P8qxVbXYneW6Sv05yWpJLu7uTfO3+BlX1kiQ3CX0AAAB3zAGDX1V9a5JfS3JNZiN+x1XV87r7dq+22d23VNXZSd6U5LAkr+ruq6rq3CTL3b07yflJLqyqPZmN9J1+57oDAADAajUbYLudBlXvTfLU7t4zTT88yeu7+ys2ob65LS0t9fLy8uZvuNY6TXGbOMB7+wX6sHgj9EMftocR+pCM0Q992B5G6EMyXz9G6EOyvfuhD9vHXWmf2EBVdUV3L621bJ6Lu3x4f+ibXJvkwxtSGQAAAAs3z8VdrqqqNyR5dWZX3PyOJJdX1bcnSXf/twXWBwAAwJ00T/C7Z5J/SvL10/S+JF+W5GmZBUHBDwAAYBub56qe37MZhQAAALAY81zV87gkP5Dk2JXtu/vpiysLAACAjTLPoZ6vzey2C6/L7D5+AAAAHELmCX6f6e5fXnglAAAALMQ8we+Xquqnk/xJks/un9ndb1tYVQAAAGyYeYLfVyV5dpJvyL8c6tnTNAAAANvcPMHv25I8rLtvXnQxAAAAbLy7zdHm7Unut+hCAAAAWIx5RvwekOS9VXV5vvgcP7dzAAAAOATME/x+euFVAAAAsDAHDH7d/ebNKAQAAIDFWDf4VdUnM7t6520WJenu/tKFVQUAAMCGWTf4dfd9N7MQAAAAFmOeq3oCAABwCBP8AAAABif4AQAADE7wAwAAGJzgBwAAMDjBDwAAYHCCHwAAwOAEPwAAgMEJfgAAAIMT/AAAAAYn+AEAAAxO8AMAABic4AcAADA4wQ8AAGBwgh8AAMDgBD8AAIDBCX4AAACDE/wAAAAGJ/gBAAAMTvADAAAYnOAHAAAwOMEPAABgcIIfAADA4AQ/AACAwQl+AAAAgxP8AAAABif4AQAADE7wAwAAGJzgBwAAMDjBDwAAYHCCHwAAwOAEPwAAgMEJfgAAAIMT/AAAAAYn+AEAAAxO8AMAABic4AcAADC4hQa/qnpKVV1dVXuq6pw1lh9eVRdPyy+rqmOn+SdV1ZXT4+1V9W2LrBMAAGBkCwt+VXVYkpcnOSXJriRnVNWuVc3OTHJjdx+f5LwkL53mvyvJUnc/JslTkvx6Ve1YVK0AAAAjW+SI30lJ9nT3td19c5KLkpy6qs2pSS6Ynl+S5OSqqu7+dHffMs2/Z5JeYJ0AAABDW2Twe0iS61ZM753mrdlmCnofT3JkklTVE6rqqiTvTPL8FUHwC6rqrKparqrlffv2LaALAAAAh75FBr9aY97qkbt123T3Zd19YpLHJ/mJqrrnbRp2v6K7l7p7aefOnXe6YAAAgBEtMvjtTXLMiumjk1y/XpvpHL4jktywskF3vyfJp5I8amGVAgAADGyRwe/yJCdU1XFVdY8kpyfZvarN7iTPnZ6fluTS7u5pnR1JUlUPTfLIJB9cYK0AAADDWtiVMrv7lqo6O8mbkhyW5FXdfVVVnZtkubt3Jzk/yYVVtSezkb7Tp9WflOScqvpcks8n+f7u/siiagUAABhZdY9xwcylpaVeXl7e/A3XWqcpbhPzvrf6sHgj9EMftocR+pCM0Q992B5G6EMyXz9G6EOyvfuhD9vHXWmf2EBVdUV3L621bKE3cAcAAGDrCX4AAACDE/wAAAAGJ/gBAAAMTvADAAAYnOAHAAAwOMEPAABgcIIfAADA4AQ/AACAwQl+AAAAgxP8AAAABif4AQAADE7wAwAAGJzgBwAAMDjBDwAAYHCCHwAAwOAEPwAAgMEJfgAAAIMT/AAAAAYn+AEAAAxO8AMAABic4AcAADA4wQ8AAGBwgh8AAMDgBD8AAIDBCX4AAACDE/wAAAAGJ/gBAAAMTvADAAAYnOAHAAAwOMEPAABgcIIfAADA4AQ/AACAwQl+AAAAgxP8AAAABif4AQAADE7wAwAAGJzgBwAAMDjBDwAAYHCCHwAAwOAEPwAAgMEJfgAAAIMT/AAAAAYn+AEAAAxO8AMAABic4AcAADA4wQ8AAGBwgh8AAMDgBD8AAIDBCX4AAACDE/wAAAAGJ/gBAAAMTvADAAAY3EKDX1U9paqurqo9VXXOGssPr6qLp+WXVdWx0/xvrKorquqd07/fsMg6AQAARraw4FdVhyV5eZJTkuxKckZV7VrV7MwkN3b38UnOS/LSaf5HkhJPK+8AABO7SURBVDytu78qyXOTXLioOgEAAEa3yBG/k5Ls6e5ru/vmJBclOXVVm1OTXDA9vyTJyVVV3f233X39NP+qJPesqsMXWCsAAMCwFhn8HpLkuhXTe6d5a7bp7luSfDzJkava/Oskf9vdn129gao6q6qWq2p53759G1Y4AADASBYZ/GqNeX0wbarqxMwO/3zeWhvo7ld091J3L+3cufMOFwoAADCyRQa/vUmOWTF9dJLr12tTVTuSHJHkhmn66CR/mOQ53X3NAusEAAAY2iKD3+VJTqiq46rqHklOT7J7VZvdmV28JUlOS3Jpd3dV3S/J65P8RHe/ZYE1AgAADG9hwW86Z+/sJG9K8p4kr+7uq6rq3Kp6+tTs/CRHVtWeJD+cZP8tH85OcnySn6qqK6fH/RdVKwAAwMiqe/Vpd4empaWlXl5e3vwN11qnKW4T8763+rB4I/RDH7aHEfqQjNEPfdgeRuhDMl8/RuhDsr37oQ/bx11pn9hAVXVFdy+ttWyhN3AHAABg6wl+AAAAgxP8AAAABif4AQAADE7wAwAAGJzgBwAAMDjBDwAAYHCCHwAAwOAEPwAAgMEJfgAAAIMT/AAAAAYn+AEAAAxO8AMAABic4AcAADA4wQ8AAGBwgh8AAMDgBD8AAIDBCX4AAACDE/wAAAAGJ/gBAAAMTvADAAAYnOAHAAAwOMEPAABgcIIfAADA4AQ/AACAwQl+AAAAgxP8AAAABif4AQAADE7wAwAAGJzgBwAAMDjBDwAAYHCCHwAAwOAEPwAAgMEJfgAAAIMT/AAAAAYn+AEAAAxO8AMAABic4AcAADA4wQ8AAGBwgh8AAMDgBD8AAIDBCX4AAACDE/wAAAAGJ/gBAAAMTvADAAAYnOAHAAAwOMEPAABgcIIfAADA4AQ/AACAwQl+AAAAgxP8AAAABif4AQAADE7wAwAAGNxCg19VPaWqrq6qPVV1zhrLD6+qi6fll1XVsdP8I6vqz6rqpqp62SJrBAAAGN3Cgl9VHZbk5UlOSbIryRlVtWtVszOT3Njdxyc5L8lLp/mfSfJTSV60qPoAAADuKhY54ndSkj3dfW1335zkoiSnrmpzapILpueXJDm5qqq7P9Xdf5lZAAQAAOBOWGTwe0iS61ZM753mrdmmu29J8vEkR867gao6q6qWq2p53759d7JcAACAMS0y+NUa8/oOtFlXd7+iu5e6e2nnzp0HVRwAAMBdxSKD394kx6yYPjrJ9eu1qaodSY5IcsMCawIAALjLWWTwuzzJCVV1XFXdI8npSXavarM7yXOn56clubS75x7xAwAA4MB2LOqFu/uWqjo7yZuSHJbkVd19VVWdm2S5u3cnOT/JhVW1J7ORvtP3r19VH0zypUnuUVXPSPJN3f3uRdULAAAwqoUFvyTp7jckecOqef9+xfPPJPmOddY9dpG1AQAA3FUs9AbuAAAAbD3BDwAAYHCCHwAAwOAEPwAAgMEJfgAAAIMT/AAAAAYn+AEAAAxO8AMAABic4AcAADA4wQ8AAGBwgh8AAMDgBD8AAIDBCX4AAACDE/wAAAAGJ/gBAAAMTvADAAAYnOAHAAAwOMEPAABgcIIfAADA4AQ/AACAwQl+AAAAgxP8AAAABif4AQAADE7wAwAAGJzgBwAAMDjBDwAAYHCCHwAAwOAEPwAAgMEJfgAAAIMT/AAAAAYn+AEAAAxO8AMAABic4AcAADA4wQ8AAGBwgh8AAMDgBD8AAIDBCX4AAACDE/wAAAAGJ/gBAAAMTvADAAAYnOAHAAAwOMEPAABgcIIfAADA4AQ/AACAwQl+AAAAgxP8AAAABif4AQAADE7wAwAAGJzgBwAAMDjBDwAAYHCCHwAAwOAEPwAAgMEJfgAAAINbaPCrqqdU1dVVtaeqzllj+eFVdfG0/LKqOnbFsp+Y5l9dVd+8yDoBAABGtrDgV1WHJXl5klOS7EpyRlXtWtXszCQ3dvfxSc5L8tJp3V1JTk9yYpKnJPmV6fUAAAA4SIsc8TspyZ7uvra7b05yUZJTV7U5NckF0/NLkpxcVTXNv6i7P9vdH0iyZ3o9AAAADtKOBb72Q5Jct2J6b5InrNemu2+pqo8nOXKa/zer1n3I6g1U1VlJzpomb6qqqzem9C11VJKPbMgrVW3Iy9wB+rDa1vRjhD4kPk9fbIQ+JPaJO26EPiT2idvyebqjRuhDYp+4LZ+nO+qh6y1YZPBbq6c9Z5t51k13vyLJKw6+tO2rqpa7e2mr67gz9GF7GKEPyRj90IftQR+2jxH6oQ/bwwh9SMbohz5sf4s81HNvkmNWTB+d5Pr12lTVjiRHJLlhznUBAACYwyKD3+VJTqiq46rqHpldrGX3qja7kzx3en5akku7u6f5p09X/TwuyQlJ3rrAWgEAAIa1sEM9p3P2zk7ypiSHJXlVd19VVecmWe7u3UnOT3JhVe3JbKTv9Gndq6rq1UneneSWJC/o7lsXVes2M8Khq/qwPYzQh2SMfujD9qAP28cI/dCH7WGEPiRj9EMftrmaDbABAAAwqoXewB0AAICtJ/gBAAAMTvDbRFV1dFX9UVW9v6quqapfmi58s7pdVdVPTu3eV1V/VlUnbkXN66mqf1dVV1XVO6rqyqp6wqrlx1ZVV9UPrJj3sqr67k0vdh1V9cCqumh6L95dVW+oqkes0e45VfWuqb/vrqoXbUW9a5n3MzW1fVFVvXfqy9ur6jmbXe96DvR5mto8YnqP9lTVe6rq1VX1gK2od7WD2Ld/q6r+oaoOn6aPqqoPbnrBa5j2119cMf2iqnrJOm2/t6reOb1f76qqUzet0AOYZ7+uqkdW1Z9Pn7X3VNW2OqfjIH43nTXt0++tqrdW1ZO2ot61zLlPn1hVl07fc++vqp+q2robkK10kPvDEVX129P7dc30/IhNK/Z2HGQ/7lNVvz714aqq+ou13rfNdpB9+GBV/cGK6dOq6rcWX+WBHcT3xBOr6rIVv59esgXlrmnO368vqapPV9X9V8y7afOrvX0H+h1VVXerql+evuPeWVWX1+yCk4c0wW+TTF9m/y3Ja7v7hCSPSHKfJD+7RvMXJPk/knx1dz8iyX9Osruq7rlZ9d6eqvrfkzw1yWO7+9FJ/lWS69Zo+uEk/3atX2xbbXo//jDJn3f3w7t7V5IXJ3nAqnanJHlhkm/q7hOTPDbJxze73rUczGeqqp6f5BuTnNTdj0rydVn7fpmbbp7P0/TZf32SX+3u47v7K5P8apKdm13vage5byfJrUm+d5PKOxifTfLtVXXU7TWqqqOT/LskT5rerycmeccm1HdA8+7XSX45yXnd/Zjps/T/bXKp6zqI301PTfK8zN6Hr0jy/CS/V1UP3OyaV5tzn75XZlfw/rnpe+6rM/ve+/5NLnc9c+0Pk/OTXDu9Xw9P8oEkv7HQ6uZ3MP34jcwutHfC9H333ZndzHqrHUwfkmSptt9/lh/M98QFSc7q7sckeVSSV29aobfjIH6/JrMboP/IZtZ3MOb8O/aZSR6c5NHd/VVJvi3Jxza10AUQ/DbPNyT5THf/ZpJMVyn9oSTfW1Vfsqrtjyf5ge7+9NT2T5L8VZLv3MR6b8+Dknykuz+bJN39ke5e6z6L+5L8af7llh3byf+V5HPd/Wv7Z3T3ld39P1e1+4kkL9rfv+7+THe/chPrvD0H85l6cZLv7+5PTG0/3t0XbGq165vn8/SsJH/d3a/bP6O7/6y737WJda7nYN6HJPmvSX6oZvcu3U5uyexqZj90gHb3T/LJJDclSXff1N0fWHBt85p3v35QZveL3d/mnZtU3zzm7cOPJ/nR7v7I1OZtmf3B+IJNq3R98+7Tb5m+3zJ9352d5JxNrXR9c+0PVXV8kscl+Q8rZp+bWfh4+OLKm9u8/Xh4kick+cnu/nySdPe13f36xZd4QPP+btrvFzL7zttODuZ74v5JPrS/XXe/e1MrXd+8v5uS5FVJnllVX7Zp1R2ceX5HPSjJh1bsD3u7+8ZNrnPDCX6b58QkV6ycMf0R/vdJjt8/r6q+NMm9u/uaVesvT6+xHfxJkmOmw3N+paq+/nba/lySH6mqwzaptnk9KqvejzvZbivM+5m6b5L7rvGZ2i7m+Twd8u/DCn+f5C+TPHvxpR20lyf5zgMcpvb2JP+U5ANV9ZtV9bTNKW0u835OzktyaVW9sap+qKrut+C6Dsa8fbjN5y7b53tinn16rf3mmiT3mb4Ht4N59oddSa5cecup6fmV2R7vRTJfP07Mqn5sM/P0Yb9XJ3nsFMq3i4P5njgvydVV9YdV9bztcrRXDu57+KbMwt+/XVw5d8o8v6NeneRp02Ggv1hVX7PJNS6E4Ld5Ksla985Yb/4dbbdw3X1TZv/DeVZmo3oX1zrn7k0jAW/N7H932Vjzfqa2zWdnLQfzedqm7si+/Z+S/Gi22e/g6Q+R307yg7fT5tYkT0lyWpL3JTlvO52DMo/pf92/Mslrkjw5yd/UdN7lIW5b7Otz7tO3V+uW9yGZb3/Inf9uX7g5+7GtHWQfbk3y/2Z2xM52MffnpLvPTbKUWTh5VpI/Xnh1i/HLSZ67jf4j5wvm+R3V3XuTPDKzz9Hnk/xpVZ28yaVuuG31R8fgrspsR/6CaWc4JskPT/+j8Ibpl9unquphq9Z/bGY3tN8WpsMP/ry7fzqzw3OeOfXhyqp6+qrm/ymzw5K20+ftqsx2+i9SVT+7vx+3126buLOfqW1jjs/TIf8+rFze3XsyGxH4vzetyvn91yRnJrl3klTVYSvei3OTpGfe2t3/OcnpSf711pX7Rebdr9Pd13f3q7r71MwOJXvUJtZ5e+btw7vXaLdtvifm3KdX7zcPS3JTd39yC0pez4H2h6uSfE1VfeH7bXr+1UnesxUFr2Oefnz1yn5sQwf83bTChZmdy/7lm1zjeg7qe6K7r+nuX01ycmbvy5GbWu3a5v79miTd/bEkv5ftc97uF5nn79ju/mx3v7G7fzSzv2WfsZU1b4ju9tiER2b/q7Oc5H+1d/+hd1d1HMefr82xhnMLsppTaCAbmVZiiiBiM00kozGYf61EmOEPcH/1R3/Y2B8q0fQvhxEl6B/94cT+WKNaJBH9oJib+/HdWGjuK4aGFNSyFCqPf5zzhY/f7va99zv3vdfPng+4fD/nfs/nft6Hz6977vmcc+5o6cXA94FHB+TdCuwBlrX0zcDLM+lxv6i/gKztpB8Eds7KswaY6qR3UR9puHPc8Xf2xx+Ar3feuwb4/Kx8X2r7bVVLLwW2jjv+eRxT9wE/BVa09Apq5/FJKMcwx9My4CXgts57twKfnoD4R9kPTwKb2vLlwDQwPe4ytHje7Cx/p52v2wfkW03tED+TvgvYM+74O/timPP6VmBJW15F7U+zatzxj1iGrwD7gI+09JVtn100AWUY9px+Gbi5k95D7d8+CfthqPOh/f9HwLZOehvw7LjLMI9y7KL2VUxLrwU2fMDKMA1c2Jbva3mfnIAyjHKfuK2zDy6jDpSyeELKMMy1aTt1bASogwOdoPZvHGv8s2Ic5hp1FbC6LS+itjh/Y9yxn+lrkn/Z6ZVSj5yNwO1JXqQ+IvU2gzsgP0a9oR9J8kfgW9SL71sLFe8clgNPpQ7le5jax2H7HOs8BFxytgMbVmd/fDFt6GpqGV6ble8n1L4Fv2h59gMTMSjHiMfUd4FfAvuSTAG/Av69ULHOYc7jqR37XwbuTx0K+xh1xLk3FjjW/zPifuiudxQ4cPYjnJdHOfVofkuAR1KnEDhIHflsIvpxDHteA7cAU0kOAXupg6T8ZUGDPYURrk27qX1ofpfkOPVL5FdLKa8vcMiDDHtObwAeaPe5I9T73s4FjnUYpzsfoLZCrUudauZP1BEbtyxIZKOZqxx3UX8IeSnJEeoxNWjgtnGaqwxdT/DBvF9/jdrH7yC15XJzmYC+lyNcX7vr/JU6EuikPUo/zPfYjwE/bt+ZDlOfDJnE69NIZn5RkCRJkiT1lC1+kiRJktRzVvwkSZIkqees+EmSJElSz1nxkyRJkqSes+InSZIkST1nxU+SdM5L8r/O5L0Hk6yZx2d8OMlETlYsSZLTOUiSznlJ3iylLD/Dz1hDncz+ihHXWzwJ83RJkvrNFj9JkgZIsjjJjiT7khxOcnd7f3mS55IcSHIkyYa2yreBS1uL4Y4k65Ps6XzeziR3tuXpJNuS/IY6qfOlSX6WZH+SXyf55EKXV5LUb+eNOwBJkibAsiQH2/KJUspGYAvwj1LKNUmWAr9N8nPgVWBjKeVkkguB3yfZDXwTuKKUciVAkvVzbPPtUsr1Le9zwD2llBeTXAs8Dnzh/S6kJOncZcVPkiR4a6bC1nEL8Jkkm1p6JbAW+DPwcJIbgHeAi4GPz2ObT0NtQQSuA55JMvO/pfP4PEmSTsmKnyRJgwW4v5Sy9z1v1sc1Pwp8rpTynyTTwIcGrP9f3tulYnaef7W/i4C/D6h4SpL0vrGPnyRJg+0F7k2yBCDJuiTnU1v+3miVvhuBT7T8/wQu6Kz/CvCpJEuTrARuGrSRUspJ4ESS29t2kuSzZ6dIkqRzlRU/SZIG+wFwDDiQZAr4HvVJmR8CVyd5HtgMHAcopfyN2g9wKsmOUsqrwC7gcFvnhdNsazOwJckh4Ciw4TR5JUkamdM5SJIkSVLP2eInSZIkST1nxU+SJEmSes6KnyRJkiT1nBU/SZIkSeo5K36SJEmS1HNW/CRJkiSp56z4SZIkSVLPvQvOZK+TzGXW6QAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 1080x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "'''\n",
    "features based on combination of:\n",
    "atom_types = ['C','N','O,'S']\n",
    "'''\n",
    "\n",
    "atom_types = ['C','N','O','S']\n",
    "feature_labels = []\n",
    "for i in range(len(atom_types)):\n",
    "    for j in range(len(atom_types)):\n",
    "        label = atom_types[i]+'-'+atom_types[j]\n",
    "        feature_labels.append(label)\n",
    "feature_labels+=['C_a_H', 'C_a_A']\n",
    "feature_labels=np.array(feature_labels)\n",
    "\n",
    "importance = rf.feature_importances_\n",
    "\n",
    "#indices = np.argsort(importance)[::-1][:20]\n",
    "indices = np.argsort(importance)[::-1]\n",
    "\n",
    "print(\"Top Feature ranking:\")\n",
    "n = importance.shape[0]\n",
    "#n = 20\n",
    "for f in range(n):\n",
    "    print(\"%d. feature %d (%f)\" % (f + 1, indices[f], importance[indices[f]]))\n",
    "\n",
    "plt.figure()\n",
    "plt.title(\"Feature importances\")\n",
    "plt.bar(range(n), importance[indices],\n",
    "       color=\"r\", align=\"center\")\n",
    "plt.xticks(range(n), feature_labels[indices])\n",
    "plt.xlim([-1, n])\n",
    "plt.xlabel('Feature')\n",
    "plt.ylabel('Importance value')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=12)]: Using backend ThreadingBackend with 12 concurrent workers.\n",
      "[Parallel(n_jobs=12)]: Done  26 tasks      | elapsed:    0.0s\n",
      "[Parallel(n_jobs=12)]: Done 176 tasks      | elapsed:    0.0s\n",
      "[Parallel(n_jobs=12)]: Done 426 tasks      | elapsed:    0.0s\n",
      "[Parallel(n_jobs=12)]: Done 776 tasks      | elapsed:    0.0s\n",
      "[Parallel(n_jobs=12)]: Done 1000 out of 1000 | elapsed:    0.0s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.177196   0.07633217 0.1437247  0.14160894 0.11939359 0.09745755\n",
      " 0.11119658 0.21882789 0.162793   0.13234326 0.24534185 0.0835425\n",
      " 0.13751559 0.36592612 0.12178895 0.17602715]\n",
      "[ 1 11  5  6  4 14  9 12  3  2  8 15  0  7 10 13]\n"
     ]
    }
   ],
   "source": [
    "'''\n",
    "permutation importance on train data\n",
    "'''\n",
    "result = permutation_importance(rf, x_train, y_train, n_repeats=10,\n",
    "                                random_state=42, n_jobs=-1)\n",
    "\n",
    "sorted_idx = result.importances_mean.argsort()\n",
    "p_importance_mean_train = result.importances_mean\n",
    "print(p_importance_mean_train)\n",
    "print(sorted_idx)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=12)]: Using backend ThreadingBackend with 12 concurrent workers.\n",
      "[Parallel(n_jobs=12)]: Done  26 tasks      | elapsed:    0.0s\n",
      "[Parallel(n_jobs=12)]: Done 176 tasks      | elapsed:    0.0s\n",
      "[Parallel(n_jobs=12)]: Done 426 tasks      | elapsed:    0.0s\n",
      "[Parallel(n_jobs=12)]: Done 776 tasks      | elapsed:    0.0s\n",
      "[Parallel(n_jobs=12)]: Done 1000 out of 1000 | elapsed:    0.0s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[ 0.02792791  0.00629706  0.01001     0.01219087 -0.00475826 -0.00480654\n",
      "  0.00537015  0.02179903  0.0247389   0.01759114  0.04163343 -0.00385844\n",
      "  0.00853628  0.07242679 -0.00254297  0.06047654]\n",
      "[ 5  4 11 14  6  1 12  2  3  9  7  8  0 10 15 13]\n"
     ]
    }
   ],
   "source": [
    "'''\n",
    "permutation importance on test data\n",
    "'''\n",
    "result = permutation_importance(rf, x_test, y_test, n_repeats=10,\n",
    "                                random_state=42, n_jobs=-1)\n",
    "\n",
    "sorted_idx = result.importances_mean.argsort()\n",
    "p_importance_mean_test = result.importances_mean\n",
    "print(p_importance_mean_test)\n",
    "print(sorted_idx)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2.5110158454401885 0.293031901471745\n"
     ]
    }
   ],
   "source": [
    "print(np.sum(p_importance_mean_train), np.sum(p_importance_mean_test))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# data to plot\n",
    "length = len(p_importance_mean_test)\n",
    "\n",
    "\n",
    "# create plot\n",
    "fig, ax = plt.subplots()\n",
    "index = np.arange(n_groups)\n",
    "bar_width = 0.35\n",
    "opacity = 0.8\n",
    "\n",
    "rects1 = plt.bar(index - bar_width/3, p_importance_mean_train, bar_width,\n",
    "alpha=opacity,\n",
    "color='b',\n",
    "label='permutation importance on training data')\n",
    "\n",
    "rects2 = plt.bar(index + bar_width/3, p_importance_mean_test, bar_width,\n",
    "alpha=opacity,\n",
    "color='g',\n",
    "label='permutation importance on testing data')\n",
    "\n",
    "plt.xlabel('Feature Index')\n",
    "plt.ylabel('Importance')\n",
    "plt.title('Permutation Feature Importance')\n",
    "plt.xticks(np.arange(0,length), feature_labels)\n",
    "plt.legend()\n",
    "\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Second type model: 18 features ([C, N, O, S] + CA_H + CA_A)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 12 Angstrom cutoff"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training Features Shape: (1812, 18)\n",
      "Training Labels Shape: (1812,)\n",
      "Testing Features Shape: (604, 18)\n",
      "Testing Labels Shape: (604,)\n"
     ]
    }
   ],
   "source": [
    "data_file = os.getcwd()+'/Data/dataset_ha_beta_12A.pkl'\n",
    "dataset = dataset_loader(data_file)\n",
    "\n",
    "features = np.array([data['x_vector'] for data in dataset])\n",
    "labels = np.array([data['y'] for data in dataset])\n",
    "x_train, x_test, y_train, y_test = train_test_split(features, labels, test_size=0.25, random_state=13)\n",
    "print('Training Features Shape:', x_train.shape)\n",
    "print('Training Labels Shape:', y_train.shape)\n",
    "print('Testing Features Shape:', x_test.shape)\n",
    "print('Testing Labels Shape:', y_test.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "'''Model Loader'''\n",
    "model_file= os.getcwd()+\"/Model/rf_pp_ha_beta_12A_split.pkl\"\n",
    "with open(model_file, \"rb\") as f:\n",
    "    rf = pickle.load(f)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=12)]: Using backend ThreadingBackend with 12 concurrent workers.\n",
      "[Parallel(n_jobs=12)]: Done  26 tasks      | elapsed:    0.0s\n",
      "[Parallel(n_jobs=12)]: Done 176 tasks      | elapsed:    0.0s\n",
      "[Parallel(n_jobs=12)]: Done 426 tasks      | elapsed:    0.0s\n",
      "[Parallel(n_jobs=12)]: Done 776 tasks      | elapsed:    0.0s\n",
      "[Parallel(n_jobs=12)]: Done 1000 out of 1000 | elapsed:    0.0s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean Absolute Error: 1.41\n",
      "Accuracy: 77.63 %.\n",
      "Root Mean Squared Error : 1.8\n",
      "Pearson Correlation Coefficient : 0.43\n"
     ]
    }
   ],
   "source": [
    "'''Model evaluation on Test set'''\n",
    "#Mean Absolute Error\n",
    "preds = rf.predict(x_test)\n",
    "errors = abs(preds - y_test)\n",
    "print('Mean Absolute Error:', round(np.mean(errors), 2))\n",
    "\n",
    "#Mean Absolute Percentage Error & Accuracy\n",
    "mape = 100 * (errors / y_test)\n",
    "accuracy = 100 - np.mean(mape)\n",
    "print('Accuracy:', round(accuracy, 2), '%.')\n",
    "\n",
    "#Root Mean Squared Error\n",
    "rmse = sqrt(mean_squared_error(y_test, preds))\n",
    "print('Root Mean Squared Error :', round(rmse, 2))\n",
    "\n",
    "#Pearson Correlation Coefficient (PCC) score\n",
    "pcc = pearsonr(y_test, preds)\n",
    "print('Pearson Correlation Coefficient :', round(pcc[0],2))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "\n",
    "fig_size = np.zeros(2)\n",
    "fig_size[0] = 15\n",
    "fig_size[1] = 10\n",
    "plt.rcParams[\"figure.figsize\"] = fig_size\n",
    "\n",
    "f, ax = plt.subplots()\n",
    "\n",
    "plt.plot(y_test, preds, 'o', markerfacecolor=\"None\", markeredgecolor='green', markeredgewidth=1)\n",
    "#ax.plot(ax.get_xlim(), ax.get_ylim(), ls=\"--\", c=\".3\")\n",
    "coef = np.polyfit(y_test,preds,1)\n",
    "poly1d_fn = np.poly1d(coef)\n",
    "plt.plot(y_test, poly1d_fn(y_test), '--k')\n",
    "\n",
    "plt.title(\"Test data binding affinity\")\n",
    "plt.xlabel('Actual binding affinity data')\n",
    "plt.ylabel('Predicted binding affinity data')\n",
    "#plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Top Feature ranking:\n",
      "1. feature 17 (0.108756)\n",
      "2. feature 16 (0.101709)\n",
      "3. feature 10 (0.075521)\n",
      "4. feature 8 (0.060101)\n",
      "5. feature 9 (0.059726)\n",
      "6. feature 0 (0.056625)\n",
      "7. feature 13 (0.053568)\n",
      "8. feature 6 (0.049845)\n",
      "9. feature 4 (0.049822)\n",
      "10. feature 2 (0.049602)\n",
      "11. feature 12 (0.047659)\n",
      "12. feature 5 (0.043194)\n",
      "13. feature 1 (0.043105)\n",
      "14. feature 7 (0.043006)\n",
      "15. feature 14 (0.042519)\n",
      "16. feature 3 (0.041040)\n",
      "17. feature 11 (0.040243)\n",
      "18. feature 15 (0.033961)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "'''\n",
    "features based on combination of:\n",
    "atom_types = ['C','N','O,'S']\n",
    "'''\n",
    "\n",
    "atom_types = ['C','N','O','S']\n",
    "feature_labels = []\n",
    "for i in range(len(atom_types)):\n",
    "    for j in range(len(atom_types)):\n",
    "        label = atom_types[i]+'-'+atom_types[j]\n",
    "        feature_labels.append(label)\n",
    "feature_labels+=['C_a_H', 'C_a_A']\n",
    "feature_labels=np.array(feature_labels)\n",
    "\n",
    "importance = rf.feature_importances_\n",
    "\n",
    "#indices = np.argsort(importance)[::-1][:20]\n",
    "indices = np.argsort(importance)[::-1]\n",
    "\n",
    "print(\"Top Feature ranking:\")\n",
    "n = importance.shape[0]\n",
    "#n = 20\n",
    "for f in range(n):\n",
    "    print(\"%d. feature %d (%f)\" % (f + 1, indices[f], importance[indices[f]]))\n",
    "\n",
    "\n",
    "plt.figure()\n",
    "plt.title(\"Feature importances\")\n",
    "plt.bar(range(n), importance[indices],\n",
    "       color=\"r\", align=\"center\")\n",
    "plt.xticks(range(n), feature_labels[indices])\n",
    "plt.xlim([-1, n])\n",
    "plt.xlabel('Feature')\n",
    "plt.ylabel('Importance value')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=12)]: Using backend ThreadingBackend with 12 concurrent workers.\n",
      "[Parallel(n_jobs=12)]: Done  26 tasks      | elapsed:    0.0s\n",
      "[Parallel(n_jobs=12)]: Done 176 tasks      | elapsed:    0.0s\n",
      "[Parallel(n_jobs=12)]: Done 426 tasks      | elapsed:    0.0s\n",
      "[Parallel(n_jobs=12)]: Done 776 tasks      | elapsed:    0.0s\n",
      "[Parallel(n_jobs=12)]: Done 1000 out of 1000 | elapsed:    0.0s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[ 0.02725345  0.01000416  0.00760272  0.0068398   0.02740422  0.00322749\n",
      "  0.00595238  0.00261261  0.02004775  0.03816556  0.03529331  0.00086465\n",
      "  0.00655367  0.04478114 -0.00309697  0.03492419  0.05808484  0.19752509]\n",
      "[17 16 13  9 10 15  4  0  8  1  2  3 12  6  5  7 11 14]\n"
     ]
    }
   ],
   "source": [
    "'''\n",
    "permutation importance on test data\n",
    "'''\n",
    "result = permutation_importance(rf, x_test, y_test, n_repeats=10,\n",
    "                                random_state=42, n_jobs=-1)\n",
    "\n",
    "sorted_idx = result.importances_mean.argsort()[::-1]\n",
    "p_importance_mean_test = result.importances_mean\n",
    "print(p_importance_mean_test)\n",
    "print(sorted_idx)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# data to plot\n",
    "length = len(sorted_idx)\n",
    "\n",
    "atom_types = ['C','N','O','S']\n",
    "feature_labels = []\n",
    "for i in range(len(atom_types)):\n",
    "    for j in range(len(atom_types)):\n",
    "        label = atom_types[i]+'-'+atom_types[j]\n",
    "        feature_labels.append(label)\n",
    "feature_labels+=['C_a_H', 'C_a_A']\n",
    "feature_labels=np.array(feature_labels)\n",
    "indices = sorted_idx\n",
    "# create plot\n",
    "\n",
    "plt.figure()\n",
    "plt.title(\"Permutation feature importances\")\n",
    "plt.bar(range(length), p_importance_mean_test[indices],\n",
    "       color=\"g\", align=\"center\")\n",
    "plt.xticks(range(length), feature_labels[indices])\n",
    "plt.xlim([-1, length])\n",
    "plt.xlabel('Feature')\n",
    "plt.ylabel('Importance value')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=12)]: Using backend ThreadingBackend with 12 concurrent workers.\n",
      "[Parallel(n_jobs=12)]: Done  26 tasks      | elapsed:    0.0s\n",
      "[Parallel(n_jobs=12)]: Done 176 tasks      | elapsed:    0.0s\n",
      "[Parallel(n_jobs=12)]: Done 426 tasks      | elapsed:    0.0s\n",
      "[Parallel(n_jobs=12)]: Done 776 tasks      | elapsed:    0.1s\n",
      "[Parallel(n_jobs=12)]: Done 1000 out of 1000 | elapsed:    0.1s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.11673993 0.06239268 0.10561461 0.06061745 0.10864211 0.06283129\n",
      " 0.0975277  0.08867516 0.13241765 0.1496225  0.20302726 0.05719417\n",
      " 0.07636657 0.18162455 0.06540092 0.08135953 0.26463167 0.56598616]\n",
      "[11  3  1  5 14 12 15  7  6  2  4  0  8  9 13 10 16 17]\n"
     ]
    }
   ],
   "source": [
    "'''\n",
    "permutation importance on train data\n",
    "'''\n",
    "result = permutation_importance(rf, x_train, y_train, n_repeats=10,\n",
    "                                random_state=42, n_jobs=-1)\n",
    "\n",
    "sorted_idx = result.importances_mean.argsort()\n",
    "p_importance_mean_train = result.importances_mean\n",
    "print(p_importance_mean_train)\n",
    "print(sorted_idx)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=12)]: Using backend ThreadingBackend with 12 concurrent workers.\n",
      "[Parallel(n_jobs=12)]: Done  26 tasks      | elapsed:    0.0s\n",
      "[Parallel(n_jobs=12)]: Done 176 tasks      | elapsed:    0.0s\n",
      "[Parallel(n_jobs=12)]: Done 426 tasks      | elapsed:    0.0s\n",
      "[Parallel(n_jobs=12)]: Done 776 tasks      | elapsed:    0.0s\n",
      "[Parallel(n_jobs=12)]: Done 1000 out of 1000 | elapsed:    0.0s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[ 0.02725345  0.01000416  0.00760272  0.0068398   0.02740422  0.00322749\n",
      "  0.00595238  0.00261261  0.02004775  0.03816556  0.03529331  0.00086465\n",
      "  0.00655367  0.04478114 -0.00309697  0.03492419  0.05808484  0.19752509]\n",
      "[14 11  7  5  6 12  3  2  1  8  0  4 15 10  9 13 16 17]\n"
     ]
    }
   ],
   "source": [
    "'''\n",
    "permutation importance on test data\n",
    "'''\n",
    "result = permutation_importance(rf, x_test, y_test, n_repeats=10,\n",
    "                                random_state=42, n_jobs=-1)\n",
    "\n",
    "sorted_idx = result.importances_mean.argsort()\n",
    "p_importance_mean_test = result.importances_mean\n",
    "print(p_importance_mean_test)\n",
    "print(sorted_idx)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# data to plot\n",
    "length = len(p_importance_mean_test)\n",
    "\n",
    "\n",
    "# create plot\n",
    "fig, ax = plt.subplots()\n",
    "index = np.arange(length)\n",
    "bar_width = 0.35\n",
    "opacity = 0.8\n",
    "\n",
    "rects1 = plt.bar(index - bar_width/2, p_importance_mean_train, bar_width,\n",
    "alpha=opacity,\n",
    "color='b',\n",
    "label='permutation importance on training data')\n",
    "\n",
    "rects2 = plt.bar(index + bar_width/2, p_importance_mean_test, bar_width,\n",
    "alpha=opacity,\n",
    "color='g',\n",
    "label='permutation importance on testing data')\n",
    "\n",
    "plt.xlabel('Feature Index')\n",
    "plt.ylabel('Importance')\n",
    "plt.title('Permutation Feature Importance')\n",
    "plt.xticks(np.arange(0,length), feature_labels)\n",
    "plt.legend()\n",
    "\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 4 Angstrom cutoff"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training Features Shape: (1812, 18)\n",
      "Training Labels Shape: (1812,)\n",
      "Testing Features Shape: (604, 18)\n",
      "Testing Labels Shape: (604,)\n"
     ]
    }
   ],
   "source": [
    "data_file = os.getcwd()+'/Data/dataset_ha_beta_4A.pkl'\n",
    "dataset = dataset_loader(data_file)\n",
    "\n",
    "features = np.array([data['x_vector'] for data in dataset])\n",
    "labels = np.array([data['y'] for data in dataset])\n",
    "x_train, x_test, y_train, y_test = train_test_split(features, labels, test_size=0.25, random_state=13)\n",
    "print('Training Features Shape:', x_train.shape)\n",
    "print('Training Labels Shape:', y_train.shape)\n",
    "print('Testing Features Shape:', x_test.shape)\n",
    "print('Testing Labels Shape:', y_test.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\beryl\\Anaconda3\\lib\\site-packages\\sklearn\\base.py:318: UserWarning: Trying to unpickle estimator DecisionTreeRegressor from version 0.21.2 when using version 0.22.1. This might lead to breaking code or invalid results. Use at your own risk.\n",
      "  UserWarning)\n",
      "C:\\Users\\beryl\\Anaconda3\\lib\\site-packages\\sklearn\\base.py:318: UserWarning: Trying to unpickle estimator RandomForestRegressor from version 0.21.2 when using version 0.22.1. This might lead to breaking code or invalid results. Use at your own risk.\n",
      "  UserWarning)\n"
     ]
    }
   ],
   "source": [
    "'''Model Loader'''\n",
    "model_file= os.getcwd()+\"/Model/rf_pp_ha_beta_4A_split.pkl\"\n",
    "with open(model_file, \"rb\") as f:\n",
    "    rf = pickle.load(f)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=12)]: Using backend ThreadingBackend with 12 concurrent workers.\n",
      "[Parallel(n_jobs=12)]: Done  26 tasks      | elapsed:    0.0s\n",
      "[Parallel(n_jobs=12)]: Done 176 tasks      | elapsed:    0.0s\n",
      "[Parallel(n_jobs=12)]: Done 426 tasks      | elapsed:    0.0s\n",
      "[Parallel(n_jobs=12)]: Done 776 tasks      | elapsed:    0.0s\n",
      "[Parallel(n_jobs=12)]: Done 1000 out of 1000 | elapsed:    0.0s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean Absolute Error: 1.43\n",
      "Accuracy: 77.07 %.\n",
      "Root Mean Squared Error : 1.87\n",
      "Pearson Correlation Coefficient : 0.35\n"
     ]
    }
   ],
   "source": [
    "'''Model evaluation on Test set'''\n",
    "#Mean Absolute Error\n",
    "preds = rf.predict(x_test)\n",
    "errors = abs(preds - y_test)\n",
    "print('Mean Absolute Error:', round(np.mean(errors), 2))\n",
    "\n",
    "#Mean Absolute Percentage Error & Accuracy\n",
    "mape = 100 * (errors / y_test)\n",
    "accuracy = 100 - np.mean(mape)\n",
    "print('Accuracy:', round(accuracy, 2), '%.')\n",
    "\n",
    "#Root Mean Squared Error\n",
    "rmse = sqrt(mean_squared_error(y_test, preds))\n",
    "print('Root Mean Squared Error :', round(rmse, 2))\n",
    "\n",
    "#Pearson Correlation Coefficient (PCC) score\n",
    "pcc = pearsonr(y_test, preds)\n",
    "print('Pearson Correlation Coefficient :', round(pcc[0],2))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA3UAAAJcCAYAAABe2o1qAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjAsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+17YcXAAAgAElEQVR4nOzdfXwU1b0/8M/Jg0mL+JhoomtY6mNCUq9Xro/h1theLfUnWrWtEau1UayWaL0Wtd176SY2ikAg/QXQ1sbW1GZta38qomiU5FZS1DYqJUi4hYIiWitQUEAwJPn+/sjOMvv8NLMzs/t5v17zCtnszJ6ZnV3OZ86Zc5SIgIiIiIiIiJwpz+oCEBERERERUeoY6oiIiIiIiByMoY6IiIiIiMjBGOqIiIiIiIgcjKGOiIiIiIjIwRjqiIiIiIiIHIyhjoiILKWU+kApVWv1aymlvqSU+kuK2y1WSolSyuX//ZdKqbvSKWuK5ThUKbVcKfWxUupX/sfmKaV2KKXeVkqdopTaleC2En4uERFZi6GOiCgHKaX26JZRpdQ+3e/T09juq0qpa40sq27bQcHJaCLykoicbtC2viUic43YVpLqARwK4EgR+aZS6mQAtwA4WUTcIvJXETkikQ2FPtfM95aIiNJTYHUBiIgo80TkUO3fSqm3AdwoIi9ZVyIyyAQA/ysiI7rfPxCRf1pYJiIiMhlb6oiIKIxSKl8p9d9KqU1Kqe1KqV8rpY7w/22cUupxpdQ/lVK7lFKvKaWOVEq1Avg3AD/3t/i1Rtl2g1Jqi1Jqm1JqVsjfzvdv7yOl1PtKqYVKKe0C5Mv+n//r3/7lSqlSf3fDbf7yPK2UKo+ze+cppdb7n/8zpVSR/7W/rJTaqCvLB0qpO5RSa/3l+bVS6hDd3z1KqX8opbYCCGrB8h+f/9JvVyn1Q38539O3hiqljtF1mXxVKTVHKRUxYCulCpRSv/e/7i6lVK9S6lT/3x4AcBeA6/3H53oAzwD4nP/3h5RSpymlhnXbe1Up9SP/z4+VUs8ppY70/y3w3EjvrVKqQynVElK+F5VS34lz/ImIyGAMdUREFMksABcBqAXgAnAAwEL/327EWE+P4wGUAJgJYEhE7gTwZ4y1+h3q/z2IUupfALQB+IZ/u27/NjQH/Ns7CsAUAJf6Xw8A/t3/81T/9p/C2P9jDwGoADDR//eFiK0ewIUATgVwhn9fo7kKwBcBnATgbADX+PfjcgC3AvgCgNMATI3zmhMAKADH+ffvIaWU1lr6MwDbABwLYAaA6+NsaymAEwGUAVgP4FEAEJG7ASwA8Kj/+DwK4KsANvl/jxa2rgEwHUA5gCMA3B76hCjv7aMArlFKKQBQSh0H4HwAv41TfiIiMhhDHRERRXIzgHtE5H0R2Q+gCcA3/BX4AwBKAZwoIsMi8mcR2Zvgdr8O4Pci8oqIfArgh9D9XyQif/Jvb0RE/gbg5xgLThGJyD9E5GkR2SciHwG4P9bz/X7i369t/ufXx3juQv9rbAPwHIB/0e3HwyKyXkT2YOz4xPIJgPtF5ICIPAlAAJyklCoGMA3Af/v3YQ2AX8fY32EReVRE9ujel7P820nVwyLyN/97+IRuH+NZ6d8PbeCZawA8z66eRESZx1BHRERB/MHtBADP+bv47QLwJsb+zzgaQAeAPwB4Qim1VSl1n1IqP8HNHwfgXe0XfxD7SPfaVf6uiP9QSn0MYDaCW/JCyzpeKfWIvzvnxwC6Yz3f713dv9/xlymaD3T//gRjg5CE7Yd/O7FsE5HRCNsqw1gL3tYo5Qvi7345398t9mOMtdQpjL0vqYq2jzGJiADoxMGup9cC+FUa5SAiohQx1BERURB/Zf09ABeKyBG6pVhEtovIpyIyW0ROw1iXyK8BuFpbPc7m/46xwAgAUEodDuBw3d8fBvAGxloBDwPQjLHQEm3b92CsG+e/+Z9/ke750Zyg+3cFgPfjPD+SoP3wbycVH2Bsv47XPXZClOcCwA0Y28c6jB230/yPx9tnI0Q6/p0ArlJKnYmxcj+bgXIQEVEIhjoiIorkIQBzlFInAIHBPC71//tL/ha1PAAfAxgGoI22+A8An4ux3d8CuEIpdbZ/gJIfA9C3YI0H8JGI7FFKTQJwk/YHf3fNj0K2Px5jrUu7lFIlAP4rgX27TSlV7n/+PQB+k8A6kfbjRjU2l9uhGGtRTJq/C+UzAJrU2JQN1fDftxfFeAD7AewAMA5jxy9Twt5bEdkEYB2AXwD4jYgMZbA8RETkx1BHRESRzAXwEoAepdRuAKsA/Kv/b8cDeBrAbgBrMXavmTY4xkIA1ymldiqlwuZpE5E3AdyJsXu3tgLYAmC77il3YCws7QGwGOGBazaA3/m7hU4DMB9j3S13AOjzlyWexwH0AtgAYMC/r0nx3xf3M4zdV7YewAvJbkPnZox159yGsXsIfQA+jfLcDv/zPsBY2fvSeN1kRXtvHwVQA3a9JCKyjBrrZUNERER2oJT6CYBiEbnZ6rIkQil1EYAlInKS1WUhIspVnHyciIjIQv4ul4KxboznArgOsUfktA3/vH23YazVkoiILMLul0RERNY6HGP31e0F8BiAH4vI89YWKT7/nIM7MXaf32KLi0NElNPY/ZKIiIiIiMjB2FJHRERERETkYI64p66kpETcbrfVxSAiIiIiIrLE66+/vl1ESiP9zRGhzu12o7+/3+piEBERERERWUIp9U60v7H7JRERERERkYMx1BERERERETkYQx0REREREZGDMdQRERERERE5GEMdERERERGRgzHUERERERERORhDHRERERERkYMx1BERERERETkYQx0REREREZGDMdQRERERERE5GEMdERERERGRgzHUERERERERORhDHRERERERkYMx1BERERERETkYQx0REREREZGDmRbqlFKPKKU+VEqt1T12lFLqRaXUBv/PI816fSIiIiIiolxgZkvdLwF8OeSxewCsEJGTAazw/05EREREREQpMi3UicjLAP4Z8vBlAB71//tRAJeb9fpERERERES5INP31B0rIn8HAP/PY6I9USk1QynVr5Tq37ZtW8YKSERERERE5CS2HShFRH4mIpNFZHJpaanVxSEiIiIiIrKlTIe6fyilygHA//PDDL8+EREREZGt+QZ8qF5SjfzmfFQvqYZvwGd1kcjmCjL8eksBXA9gjv/n0xl+fSIiIiIi2/IN+ODp8aBjWgdqK2rRt6UPDUsbAAD1NfUWl47syswpDXwAXgFwqlJqq1KqAWNh7j+UUhsA/If/dyIiIiIiAtCysgUd0zpQN7EOhfmFqJtYh45pHWhZ2WJ10cjGTGupE5FolxK+aNZrEhERERE52eD2QdRW1AY9VltRi8HtgxaViJzAtgOlEBERERHlmsqSSvRt6Qt6rG9LHypLKi0qETkBQx0RERERkU14pnjQsLQBvZt7cWDkAHo396JhaQM8UzxWF41sLNMDpRARERERURTaYCiNyxsxuH0QlSWVaLmwhYOkUExKRKwuQ1yTJ0+W/v5+q4tBRERERERkCaXU6yIyOdLf2P2SiIiIiIjIwRjqiIiIiIiIHIyhjoiIiChL+AZ8qF5SjfzmfFQvqYZvwGd1kYgoAzhQChEREVEW8A344OnxoGNaB2oratG3pQ8NSxsAgINsEGU5ttQRERERZYGWlS3omNaBuol1KMwvRN3EOnRM60DLyhari0ZEJmOoIyIiIsoCg9sHUVtRG/RYbUUtBrcPWlQiIsoUhjoiIiKiLFBZUom+LX1Bj/Vt6UNlSaVFJSKiTGGoIyKilHFQBiL78EzxoGFpA3o39+LAyAH0bu5Fw9IGeKZ4rC4aEZmMA6UQEVFKOCgDkb1on7vG5Y0Y3D6IypJKtFzYws8jUQ5QImJ1GeKaPHmy9Pf3W10MIiLSqV5Sjfap7aibWBd4rHdzLxqXN2LtrWstLBkREVH2UUq9LiKTI/2N3S+JiCglHJSBiIjIHhjqMoz3nxBRtuCgDERERPbAUJdB2v0n7VPbsd+zH+1T2+Hp8TDYEZEjcVAGIiIie+BAKRmknxQUQGBS0MbljbyJmYgch4MyEBER2QNDXQbx/hMiyjb1NfUMcURERBZj98sM4v0nRERERERkNIa6DOL9J0REREREZDR2v8wg3n9CRERERERG4+TjRERERERENsfJx4mIiIiIiLIUQx0REREREZGDMdQRERERERE5GEMdERERERGRgzHUERERERERORhDHRERERERkYMx1BERERERETkYQx0REREREZGDMdQRERERERE5GEMdERERERGRgzHUERERERERORhDHRERERERkYMx1BERERERETkYQx0REREREZGDMdQRERERERE5GEMdERERERGRgzHUERERERERORhDHRERERERkYMx1BERERERETkYQx0REREREZGDMdQRERERERE5GEMdERERERGRgzHUERERERERORhDHRERERERkYMx1BERERERmcA34EP1kmrkN+ejekk1fAM+q4tEWarA6gIQEREREWUb34APnh4POqZ1oLaiFn1b+tCwtAEAUF9Tb3HpKNuwpY6IiIiIyGAtK1vQMa0DdRPrUJhfiLqJdeiY1oGWlS1WF42yEEMdEREREZHBBrcPoraiNuix2opaDG4ftKhElM0Y6oiIiIiIDFZZUom+LX1Bj/Vt6UNlSaVFJaJsxlBHRERERGQwzxQPGpY2oHdzLw6MHEDv5l40LG2AZ4rH6qJRFuJAKURE5Ei+AR9aVrZgcPsgKksq4Zni4eADRGQb2vdR4/LGwPdUy4Ut/J4iUzDUERGR43BUOSJygvqaen4nUUaw+yURETkOR5UjIiI6iKGOiIgch6PKERERHcRQR0REjsNR5YiIiA5iqCMiIsfhqHJEREQHcaAUIiJyHI4qR0REdJASEavLENfkyZOlv7/f6mIQERERERFZQin1uohMjvQ3dr8kIiIiIiJyMIY6IiIiC/gGfKheUo385nxUL6mGb8BndZGIiMiheE8dERFRhnHydCIiMhJb6oiIiDKMk6cTEZGRGOqIiIgyjJOnExGRkRjqiIiIMoyTpxMRkZEY6oiIiDKMk6cTEZGROFAKERFRFL4BH1pWtgQmOPdM8RgykAknTyciIiMx1BEREUVg9giV9TX1DHFERGQIdr8kIiKKgCNUEhGRUzDUERERRcARKomIyCkY6oiIiCLgCJVEROQUDHVEREQRcIRKIiJyCg6UQkREFAFHqCQiIqdQImJ1GeKaPHmy9Pf3W10MIiIiIiIiSyilXheRyZH+xu6XREREREREDsZQR0RERERE5GAMdUQUkW/Ah+ol1chvzkf1kmr4BnxWF4mIiIiIImCoI6IwvgEfPD0etE9tx37PfrRPbYenx8NgR+RAvEBDRJT9GOqIKEzLyhZ0TOtA3cQ6FOYXom5iHTqmdaBlZYvVRSPKOemEMl6gISLKDQx1RBRmcPsgaitqgx6rrajF4PZBi0pElJvSDWW8QENElBssCXVKqduVUmuVUm8ppb5nRRmIKLrKkkr0bekLeqxvSx8qSyotKhFRbko3lPECDRFRbsh4qFNKVQO4CcBZAE4H8H+UUidnuhxEFJ1nigcNSxvQu7kXB0YOoHdzLxqWNsAzxWN10YhySrqhjBdoiIhyQ4EFr1kJ4FUR+QQAlFJ/APBVAHMtKAsRRVBfUw8AaFzeiMHtg6gsqUTLhS2Bx4koM7RQVjexLvBYMqFMu0DTMa0DtRW16NvSh4alDWi5kN0viYiyiRWhbi2AFqXU0QD2AfgKgP7QJymlZgCYAQAVFRUZLSARjQU7hjgia6UbyniBhogoNygRyfyLKtUA4LsA9gBYB2CfiNwR7fmTJ0+W/v6w3EdERJT1fAM+tKxsCYQyzxSPaaEsk69FRETJUUq9LiKTI/3NipY6iEgHgA4AUErdB2CrFeUgIiKyu0y1mmsjbYa2CmplICIi+7Jq9Mtj/D8rAFwBgBPmEBERWYjTHxAROZclLXUAfu+/p+4AgO+KyE6LykFERETg9AdERE5mVffLKVa8LhEREUWW7kibRERkHUu6XxIREZG9cH5KIiLnsqr7JREREdkIpz8gInIuS6Y0SBanNCCyHw59TkRERJQ5tpvSgIicjUOfExEREdkH76kjoqRx6HOyO9+AD9VLqpHfnI/qJdXwDXDmHCIiyl5sqSOipHHoc7IztiQTEVGuYUsdESVNG/pcj0OfZw+nt3KxJZmIiHINQx0RJY1Dn2cvrZWrfWo79nv2o31qOzw9HkcFO7YkExFRrmH3SyJKGoc+z176Vi4AgVauxuWNjnl/OYk2ERHlGoY6IkpJfU29Yyr5lLhsaOXSWpJD76lruZDdL4mIKDsx1BERUUA2tHKxJZmIiHINQx0REQVkSysXW5KJiCiXMNQREVEAW7mIiIicR4mI1WWIa/LkydLf3291MYiIiIiIiCyhlHpdRCZH+hunNCAiIiIiInIwhjoiIiIiIiIHY6gjIiIiIiJyMIY6IiIiIiIiB2OoIyIiIkv4BnyoXlKN/OZ8VC+phm/AZ3WRiIgciVMaEBERUcb5Bnzw9HjC5kQEwCk0iIiSxJY6IiIig7EFKr6WlS3omNaBuol1KMwvRN3EOnRM60DLSmdNdE9EZAcMdURERAbSWqDap7Zjv2c/2qe2w9PjsW2wsyqADm4fRG1FbdBjtRW1GNw+mJHXJyLKJgx1RORIbAkhu3JSC5SVAbSypBJ9W/qCHuvb0ofKkkrTXzsRufodk6v7TeR0vKeOiByH9+JQpvkGfGhZ2YLB7YOoLKmEZ4on6rnmpBaolpUtuKbmGjQubwzs2zU116BlZYvpnyXPFA8aljaEfY5bLrQ+/Obqd0yu7jdRNmCoIyLH0beEaJXtdz56Bzc8fQMAVj7IWMlWdLUWqLqJdYHH7NQCpbdu2zp8cuCTsH17e9fbpr+2duz0gbLlQvPDZCL03zEAAq2tjcsbbVE+s+TqfhNlA3a/JCLH0VpC9F3Hdt+zG0MjQ7a+d4mcKdnulFoLVO/mXhwYOYDezb1oWNoAzxRP2HOt7up2SP4hmHnWzKB9m3nWTBySf0hGXr++ph5rb12LkdkjWHvrWtsEBye1thopV/ebKBuwpY6IHEdrCdFXtns396KqtArtU9t5VZkMlWxFN9EWKDt0dRsaGUL7a+04o+yMQBnaX2vH0MhQRl7frpzU2mqkXN1vomzAljoichytJWRw+yDOPv7soJYQXlUmo6UyoEciLVB2GFClqrQK0z8/HY3LG1HcUozG5Y2Y/vnpqCqtylgZ7CiZ1tZskqv7TZQN2FJHRI6jVZBvePoGHHr/oagqrQq0hPRu7uVVZTJUqgN6xBtcxQ5d3TxTPBFbC+0wWImV7Hy/n5lydb+JsgFDHRE5klbJ0O6pq62oDVxVzvUKKRkrlYpuIl0r7dDVzehKfDKjhNpdfU29Y8uejlzdbyKnUyJidRnimjx5svT391tdDCKyoWyqRFL2qF5Sjfap7UGBrXdzLxqXN2LtrWsBRA9+VrWMpPtZstv+xMPvDiJyGqXU6yIyOdLf2FJHRI7Gq8pkR4l0rbRTVzcjBm1x0nD4dhikhojISGypIyIiMlgiLXV2YkR585vzsd+zH4X5hYHHDowcQHFLMUZmjxhe5nQ47f0hIgJit9Rx9EsiIiKDOW0UQSMGbUlllFCr2GGQGiIiIzHUERERGay+ph4tF7YETRVg13vLAGMCmZOCrJMCKBFRInhPHRERkQmcdL9nqtM26NnpHsF4jNhfIiI7YagjIiLTcIRBZzAqkDklyDopgBIRJYIDpRARkSmcNsQ9ERGRnXGgFCIiyjj9EPeF+YWBIe5bVrKLW67zDfhQvaQa+c35qF5SDd+Az+oiERE5GrtfEhGRKTjCIEXCOeKIiIzHljoiIjJFtowwyFYlY7EFl4jIeGypIyIiU+hHGNz68VbM7p2Ntz96G67xLvgGfLZtldEP7nLcocfh05FP8ZurfsNWJYOwBZeIyHgMdUREZAot9Fz31HXY+vFWuI9wo/PyTrgOc9k2GIV2DTxl0SkYHh3GB3s+CGpValzeaLuyO4XWgls3sS7wmBNbcImI7ITdL4mIyDT1NfU4vOhw9FzXg823b8Y3T/+mrbvbhXYN3PLRFnR+tTOorGxVSo+TJiknInIKttQREZGp7N7dTt/dclRGsfXjrYG/aa1H+rKyVSk9nCOOiMh4bKkjIiJT2XnAFK27ZfvUduz37If7cDfufunuwGAonikeXP/U9ag4rIKtSgaqr6nH2lvXYmT2CNbeupaBzoY4QBCRszDUERHlOLMrb3bubhfa3fK+L94HEcEPe36IAyMHUHZoGYZHhzEswyhuKUbj8kZDW5VYcSY7Cr3Y0T61HZ4eD89PylqvvfYa7rjjDsyYMQMvvPCC1cVJCbtfEhGlSN9tr7KkEp4pHse1OGRizjA7d7cL7RpaX1OP4dFhXPfUdShuKUZlSSV+8uWfmFJWztdmvmz4jFpBf7EDAAcIoqzz7rvvYvny5bj55puDHj/88MNxxhlnWFSqNImI7ZczzzxTiIjspGtNl0xsmyg9m3pkaHhIejb1yMS2idK1psvqoiVl0uJJ0rOpJ+ixnk09MmnxJItKNKZrTZdMWjxJ8pryZNLiSaYdVyv3367HPltky2fUCnlNeTI0PBT02NDwkOQ15VlUIqLUjY6OSk9Pj1x88cXyrW99S2bNmiX19fUCILBUVFRIR0eHDA8PW13cmAD0S5S8ZHlgS2RhqCMiu8mWCrkdK2+ZrIxbWfG347HPJtnyGbUCjx053fvvvy8dHR1BwU1bbrnlFhkYGJC//OUvMjIyYnVRkxIr1PGeOiKiFBg5oqOV91XZcRCT0PvczJwCob6mHi0XtqBxeaMp98zFYsdjn2lmnvvJfkZ5f+NBdr4PliiUiGD58uU455xzcP7556OmpgY9PT1oaGgIPOe0006Dz+fD8PAwlixZgurqanz+859HXl4WRaFoac9OC1vqiMhujLqSHdpS5FnhkcLmQlFeZWq3w2ivb4cuarnSgmXHY59JZu9/Ip9RrZuv8iopbC4UzwpPTr4XkWSqCzRRskZHR+Wjjz6Shx56SJRSYS1xX//612Xz5s3y1ltvyejoqNXFNRTY/ZKIyFhGVUj1FU9tm62rWqVqUVXGKpZ2q7zlUtevZI693d6ndJn9Psf7jOr/XrW4SlpXtQb9PVvPOSKnGRkZkSeeeEJqamqksLBQjj32WFm6dKnk5eUFgtwZZ5whTz/9dNaFuFAMdUSUMdlW8YzFiH3Vt0pplVx9q1QuVixzvQUrkmw8JplokY31GdWHSq0s+s9bNrYOEznByMiI7NmzR37+85/L6aefHtYSN336dOnv75d33nkn60NcKIY6IsqIbKx4mo0Vy8jMuDiQ6DbtcmFCX46ie4vEs8IT9HenB36rW2R5QYXIHg4cOCCdnZ1y4oknCgA5+uij5Y477pAvfvGLUl5eLgDk/PPPl+7u7pwLcaEY6ogoI6yupDkRu4BlRqIXHFK5MJFOCIy2bmg5lFeFlcPpgd/qi0B26PpMlIuGh4fl3XfflWXLlslll10W1hJ3zTXXyDPPPCO7d++2uqi2w1BHRBmRKwNcGI2DNZhHO7bwQtxt7qBjGSkwJ3thIp1gEmvd0HJMWjxJWle1BpUjGwK/la2iVg9SRJQrPv30U3nooYfkuOOOCwS3E088UX7yk5/I5z73OQEgX/ziF+Xll1+2uqi2x1BHRBnBlrr02aXrXzbQV9qVV0n3xu6gwBXpgkOyFybSOedjrRtajq41XeJe6A7qosvAnz5+3oiMd+DAAfnTn/4kAwMD8pWvfCWsJe7000+Xxx57zPYTfdtRrFBXkIlpE4goN2hzG3VM60BtRS36tvShYWkDWi40fn6xbFVfU5+ROdJygX6+u6rSKhTkFaBjWgcalzeivqY+4pxw2txxdRPrAo/pn+cb8KFlZQsGtw+isqQS67atS3m+wljzqIWWo76mHm9tewvzV81HcUsxKksqMzafXjbj540offv27cPPfvYzNDU1YefOnYHHn3vuOfztb3/D6aefjgkTJuC///u/MXnyZAtLmt2yaMY9IrKalRM5U+KMmmTZ7pM160OTdsFheHQYg9sHo06mHGvSZd+AD54eD9qntmO/Zz/ap7ajIK8ATX9oCtpGohOIx5p8PFI5uga68IvLfoGR2SNYe+tafq6IyBJDQ0P44x//iM2bN2Pq1Kn47Gc/i+9973uBQFdUVITHH38c559/PtavX4/Vq1fj6aefZqAzW7QmPDst7H5JRGQMowansHqQi3hli3QfndaFEV6kNPplpO6S2n1YRt9TF6scRESZtHfvXrnvvvukqKgoqBvl/fffL1dddZV885vflG984xsyMDBgdVGzHmJ0v1Rjf7e3yZMnS39/v9XFICJyvOol1Wif2h7UvbB3cy8alzdi7a1rM76dVIV2g/RM8aC+pj7QmtYxrQNbP96Ku168C0opPPClB+A6zBXoDpxKK1d+cz72e/ajML8w8NiBkQMo+nERqkqrwsqSzn4QEVllz549eP311zEyMoJnnnkGbW1tYc/5/e9/jy984Qs4+uijLShh7lJKvS4iEZs8eU8dEVEOiXUflxXbSYU+uOnv3QSC76MDgIK8AvxwxQ9x3VPXYVLppLS6A0e7366qtAqeKZ5AOGtZOXYPaSKvE+ueLgY+IsqEffv2Yf78+Zg9e3bQ4+eeey6uv/56fO9734NSCrfccgtOPvlki0pJ8TDUERHlkHgDgWR6O6kIDW51E+sCA6CEhs36mnpcVXUViluK025BjDYQ0CUnXxI1ZKYawmIFVwY7IkrH9u3b8dprr2HDhg0oLS3Ft7/9bQwNDQU95xe/+AUuu+wyHHnkkRaVkpIWrV+mnRbeU0dERuA9StlxT12saQfMnlYj0jlkxmtyehAiMsru3btl/vz5YVMLFBQUyMqVK2XWrFmyaNEi2bJli9VFpTjAKQ2IKNex5WOMtq9aq1aqQ+MbtZ1UxGolNHtajUjdJa998lrDu6Ja2b2ViJztvffew/PPP48NGzbgksVwL0wAACAASURBVEsuwYUXXojh4eGg58ybNw8zZszAYYcdhtra2ihbIidhqCOinBCry14uhTrAuLm5rJrjK1ZwsyJsmtEV1crurXbG+wyJwn300Uf43e9+h5tuuino8YKCApx77rmYPXs2qqurce6556KsrMyiUpLpojXh2Wlh90siSlesLnu5Jhu6oZq1D6ls14yuqHaeMsIqqRyTbDjXiUJt3rxZZs+eLXfeeaf4fD6prKwM61rZ1NQke/bssbqoZDDE6H5peWBLZGGoI6J08R6lMWYFkGyoOKdzbMw4BtlyXI2S7GeYwZiyxY4dO+TJJ58MC26FhYXywx/+UJ5++mnp6emRf/7zn1YXlUzGUEdEhnNahZMVvDFGh9tsOq4M/vaWbGt7rPfTad9flFvWr18v1157rdx8881y0003yaxZs4LCXH5+vjQ3N8vevXutLiplWKxQx3vqiChpThx0xMqBPezE6AE4suleRQ5OYm/J3mcY7f1ct22d476/KLt99NFHePHFF3H11VdjZGQk6G+XXHIJvv/97+PSSy/Fv/3bv6G4uNiiUpLd5VldACJyHn1FvjC/MFCR1yZdtqv6mnqsvXUtRmaPYO2ta3OyAqdVjPXSGYAjm4KQ0cdGzzfgQ/WSauQ356N6STV8A76Yjxu1/WyiDZDTu7kXB0YOoHdzLxqWNsAzxRPx+dHez0PyD3Hk9xdlj9WrV+PKK6/ElClTUF1djZdffhlf+9rXAoHu0EMPxbx587Bnzx4sW7YMp5xyCqZMmcJAR7FFa8Kz08Lul0T2wkFHnMvo7pLZ1GXRrK6k0bY789mZjp8zMNOS6TYZ7bgor+L3F2XU0NCQdHV1yamnnhp2X9yXv/xlWbdunbzyyisyNDQUf2OU08B76ojISNlUkbczO43wGGtb2RQozDjm0T4vRfcWGfI5irZ91wJXzt83lqnJ4on0Xn31VZk6daoUFxfLySefLI8++qgcddRRgSBXWloqixYtkn379lldVHIYhjoiMlS2VeQzIdmw4KRjzEEnYovWsg0vDGkxirT9ztWdAi8ccf5kmpM+W+QM+/btk6eeekquuuqqsJa4Sy+9VF544QXZuHGjHDhwwOqiksMx1BGR4RKtyLPCn1olMldbE5xyviRTTita6twL3eJuc6e97WzllPOM7Km3t1cuuOACASBut1u+9rWvyfXXXy9HHnmkAJAJEybII488wu6UZDiGOiKyBK+Ij0kloOXifYtOOV+SLacV99TBC+lc3Rn0PCecPwxbZDejo6Py97//XV555RW57rrrwlrizjnnHHnkkUdk165dMjIyYnVxKcsx1BGRJXK1tSlUKgHNiccu3Qq5U/Y5lXJGOzZGhZjQ7bhaXY44lnpmhnqGRUrU6OioPPvss3LWWWcFgtu4cePk8ccfl9LSUgEgp512mvh8PhkeHra6uJRjGOqIyBK52NoUSaohIFrrjh0rp0ZUyJ1yvphdTiMCiFNaPfVCPydda7rE3eYWeJF24HXasaDMGR0dlfXr18vf/vY3qa+vD2uJKy4uliVLlsju3btldHTU6uJSjmOoIyJLOKXlxWypVipDK/fJdtfLZOuEEe+1VedLssfJzHIaGUCc1jqlD8vaceje2C3Kq9I6DvweIr2RkRF54oknpKamJii8PfnkkzJhwgSpqamRM844Q5566imGOLIdhjoisgSvkB9kRAU7mcpp15ouKZ1bKu6FblFeJe6FbimdW2rasTei9cqK8yWV14y3TjrvdS4HEP2+a//W73uqx8EpLcBkjpGRERkYGJC//vWv8u1vfzusJQ6APPzww/LBBx9YXVSiuAwJdQCOAVChLYmuZ8TCUEfkXE5rLbCzZCqnrgUuKZtXFhQ8yuaViWuBK+y5mQ6csWT6fEm23Fr5lFdJ0b1Forwq7B65dIJpNgeQeO+t/tjlNeVJ98buoGOX6nHI5aCci0ZGRqSzs1NOOumkoOB26623yo033ijXXHON1NbWSnd3N1viyHHSCnUApgHYAGAvgM0ARgG8FW89IxeGOqLMYxizn2Qqp/BCujd2Bz3WvbFb4EXQY0a1jtmlVTbZ8zaZEJXIPqYbILI1gCR6fmjvH7wQ90J30N9TPQ52OTfJHENDQ/LGG2/I8uXLpbW1NTCtgH7p6OiQd955x+qiEqUt3VD3FwBHA3jT/3sdgJ/FWy/ONu8A8BaAtQB8AIpjPZ+hjiiz0qkEMQyaJ5n3JdFQZ2SIsPq9N3s+wESem25LW7YGkFRaRI08Dlafm2ScAwcOyC9/+UspLy8PCm4nnHCC/OY3v5Fvf/vbctNNN8nKlSutLiqR4dINdf1yMNzl+f/9p3jrxdje8f4Wv8/4f/8tgG/FWoehjiizUq3oZ2uF1E4SrZy6Wl1SPr886L0on18urtbg7pfZ1N3PyFFGIx3XRI6VESE5GwNIKueZ2dNAkDPs379f/vCHP0hra6usWLFCDjvssLCWuObmZnn33XetLiplAbt/v6Qb6l4CcCiAdn+r2k8ArIq3XoztHQ/gXQBHASgAsAzARbHWYagjyqxUK/rZ2nXMiQIDpbS5Ja8pT9xtkQdKyaYQkuh5G2lU0UTKn8ix4oWNyIy855LHN7sNDQ1JV1eXHHHEEWHh7cUXX5QZM2bInDlz5M9//rPVRaUs44Tvl3RD3TgAef4Adj2A2wAcFW+9ONu8HcAeANsA/DrKc2YA6AfQX1FRYfYxInI0oyvVqVbAsqnVx2kinQOJnBfp/idmp/8E4523XWu6xNXqCtyv1bm6M+muxcncF2Z2yLVLmE6EUecJLxxlnz179shvf/tbaWlpkVdffTViS9xtt90mf//7360uKmU5J3y/pBvqHkjksUQXAEcC6AFQCqAQwFMAro21DlvqiKIzo1Kd6jZjfSE6pQLqlHLqpXsOzHx2phTdWyTwQoruLZKZz85M+LUz+Z9grPema02XuBb4A1tbeGDTjpG7zS3dG7uD/pZMee1yftgpTCfKiGPHC0fOt3//fnnuueekqKgoLLy1t7fLnDlz5LHHHpO1a9daXVTKMU74fkk31L0R4bE18daLsb2vAejQ/X4dgCWx1mGoI4rOrEp1KhWwaBXNZCfNtooTK8oi6Z0DThmCP1Y59X/rXN0p7oVugRfianUF9kM7RvryasfIbv9ph4r0WXTCFWUzpDoFhdUhPJft3LlT5s+fL16vVx588EG5+OKLw8LcDTfcINu2bbO6qJTjnPC9mlKoA3ALgAH/VAZrdMtmAI9FWy/eAuBs/8iXnwWgADwKoDHWOgx1RNHZ7cqSGRXQTFXMnPCFHkk654BThuCP9TrJjEqpf652jOz8HkcLs8qrbPW5z5RkLkI49SKN033yySfS09Mj48ePDwtv06dPlxdeeEGeeuop2bJli9VFJQrihO+MVEPd4QDc/sFRJuiWtO6n82+7CcB6jE1p8CsARbGez1BHFJ0Tgkg6oSOT0ytkIiCbEVDTOQecMgR/rHImMyqlvrzdG7vFvdBtu/+09aK9t0X3Ftn+c2+WRD9DTvhuzAYffvih3HPPPfLd735Xrr76amlrawsLczfddBPviSNHsHvrflrdL+VgEDsGQIW2JLqeEQtDHVF0TriylE7lKpPTK5hdCTTrvUpnu04Z/TLdlrqwLppt/i6aC1y2+qyEihZYlVc5cn8yyW69GLLFvn375Pnnn5fTTjstLLydddZZsn79elm2bJn885//tLqoRFkn3XvqLgWwwd8NczOAUQBvxVvPyIWhjig2u19ZSid0ZHJ6BbMDciKjM6b6PqY62IlZA+0YfT4mek+dHUalNJK+hVEru3uhOxDe0hnNM9uxpc4Y7733ntx2221ywQUXyDnnnCPPPfdcWJi7/fbb5f3337e6qERZL91Q9xcARwN40/97HYCfxVvPyIWhjsj5Uq1QZ3p6BbMCyaTFkwReSNWiqqBtamVKt5tputMSGLXPZgbjeKNfOi2wJaJrzdh8g2XzyqR7Y7d0b+yW8vnlgTkHGVyis3MvBjufr6Ojo/Lcc8/JJZdcEhbezjvvPHn11VdlxYoVsnv3bquLSpRz0g11/XIw3OX5//2neOsZuTDUEeUuM6ZXyCR9+asWV0nrqtag8ifThTAaO+1r0b1FYRVVu4UMO1eoI3G1ugKTyGvl1Y4puxjGZsf32m5hc9OmTXLjjTfKYYcdJmeeeabMnTtXqqurg8LcPffcIx9++KEl5SOig9INdS8BOBRAu3/QlJ8AWBVvPSMXhjqi3Gbk9Apad71MVfT0gUsrU+uqVqlaVBVUpnQq53ao2Gv7prxK9n66N2jfYpUl061vdqtQJyLW+2uXQE+Js/o9Gxoakpdfflluu+22sJa4KVOmyGOPPSYbNmyQTz75JCPlIaLEpRvqxgHIB1AA4HoAtwE4Ot56Ri4MdUTmseOVbKNE2rdMV+pDK+Rda7qkanGVwIug4+30ljqtDPqy6H+PVBYj7pNLtZx6dg9BscrsxJCa6zJ9EWZwcFCuvfZaASAnnXSSnH/++eLxeKS4uFgASF5enjQ1NcnOnTtNeX0iMo4ho19auTDUEZkjFyuEma7UJ/p6Vt5TZwStoqovy95P90peU17UsqQ7omU65dSze3fFeO9vNl+YyUZmfwft2LFDBgcH5bvf/W5YS9xRRx0lc+bMkV27dsn+/fsNeT0iypxU56nbDeDjaEu09cxYGOqIzOHEVot0ZbpSn0zgSqdybnXFPrSb6aTFk0R5lRTdWxS1LOnOPZduOTVOOOetfn/JOEZfhFm9erVcccUVgeCmlJJly5bJuHHjBIAceuihMm/ePNmzZ4/Be0JEmZZu98tmALcCGA/gMAC3ALgr3npGLgx1ROZwYqtFumINEW8WO1bIjS6T0fMCmhW+7NCqSZTO5++9996T999/X26++eawljgAcv/998uHH34oQ0ND8TdGRI6Sbqh7LZHHzFwY6ojM4dRWi3TEGyLezowKYmYFm2TLZ8U9dfpyai2JyqtsE7aJQr3yyivy5S9/OSi4/fznP5dTTz1Vampq5JhjjpFFixbJvn37rC4qEZks3VC3CsB0/2Apef5/c/RLoiyQq60WsYaIt1q0YGTke+Va4BL3QnvsvxWjX2qBrrC5UDwrPDl17pP9vfPOO7Ju3TqZPXu2jB8/Pqwlbv78+bJhwwYZHR21uqhElGHphjo3gKcBbAewDcBTANzx1jNyYagjMo8duwaaza7dTmMFN6NaVbvWdAm8kO6N3UGv0bm6M2z/s+3cCJozcFH0OQMpNdl2vmTC6Oio9PT0yBe+8IWg4HbZZZfJrFmz5LLLLpPq6mrp6Ohgd0oi4uiXRER6du12GqtcRgXRSYsnibvNHfQ6PZt6xL3QHbT/2diKqz++2vHUv+92CPZOlY3nixlGR0dlw4YNsmzZMunq6pITTjghrCWuqalJ1q5da3VRiciGYoW6PBAR5RjPFA8aljagd3MvDowcQO/mXjQsbYBniiej5fAN+FC9pBr5zfmoXlKNddvWobaiNug5tRW1GNw+iMqSSvRt6Qv6W9+WPlSWVCb1moPbB9F8QXPQ/g+PDuPtj94O2v+WlS3omNaBuol1KMwvRN3EOnRM60DLypbUd9hig9sHA8dXO55bP96Kt7a9hfzmfJyy6BQcd+hxFpfS3kLPWd+AD0B2ni9GEBG89NJLOOuss6CUQl5eHk4++WRceeWVOPzww3H22WfjmmuuweOPP47h4WGICGbPno1JkyZZXXQicppoac9OC1vqiMiM0Rqt7CoWqWVDu8dLL91JpkP307XAFT76Z5tbXK3Bo3/atYtqOkKnXSidWypHzTlKJiycIN0bu6VsXpkjBsyJxczzOtY5mI3nSypGR0flzTfflIceekhee+01KS8vD2uJa2hokLfeeov3xBFR0pDmPXX58Z5j9sJQR5TbsrFrV6Sulp4VHilsLjRskulIx610bqmUzi2NeyxDy9e1pkvcbW6BFzJp8SSZ+exMx90/FXo8SuaWSEFTQdDol3bohpsqsz8nVkxBYXejo6PyP//zP1JdXR0W3p588km59tpr5Uc/+pE8/fTTDHFElLZ0Q91mAPMAVMV7rlkLQx1RbktlYmu7i9ayoQUMI8JStIq2q9UV9zX0AaFzdaeUzy+Xsnll0rm6UzwrPFLQXODIkSP1wRheSOfqzqC/O7l1yexgFas1LhsvvEQyPDwsL774orS3t8sf/vAHOe2008LC3KWXXip//etfGeKIyHDphrrxAG7C2NQGrwKYAeCweOsZuTDUUTawuruf0TK5P1plUl9x3PvpXslrynNsxTFWBdyoY5tulzitHPBC3AvdgXJMWjxJWle1BoUFJ7bKOKV1KdHzwewukPGOV7Z9x4mIjIyMyBtvvCEnnnhiWHi78847pa2tTR5++GF54YUXGOKIyHRphbqgJwP/DuA9AHsBPArgpGTWT3VhqCOny7ar2JneH60yqa9U6n+3WyU8nq41XeJqdQXCUufqzsAxnPnsTMOObaKhJV5lPDQs5DXlBUK1xoktXE74XEYrY6Tur2aH1GjdeV2trqwJckNDQ/LII4/IggULZP78+TJjxoywMFdXVyebNm2yuqhElIPSvqcOwDQATwJ4E8B/AjgWwFUA/hpvfSMWhjpyOqe0CCQq0/ujVSaVV8neT/cGVb7TCRNWtCyEdmvU7lNzLXAZXjFPJLQk8pzQMsVqqXNaa0288lq9P8nce2nkBYFo9MfDtcCV0P2ZdjY8PCz9/f3yr//6r2Hh7bzzzpOVK1dKZ2enrFmzxuqiEhGlHeo2AegAcF6Ev/3feOsbsTDUkdNl28hwVuxP15ouKbq3KKxybWbgMUO80Gb0sY0XShIJkaHHKto9dZkIFYnsk5Gvk8j+JFKeVMsc6XyoWlQVdj5YEaqdeLFq37590t7eLrfffrtcfvnl8vjjj4eFualTp3KeOCKypXRDXW2Ex86Pt56RC0Md2UkqlSYnVn5isWp/jAxiVu1DvNCW6XIlGiJDz3sruv9p5chUGE8l8KbaGppMGZRXyfGtxwcd/87VnRm/SOSEi1XDw8PyyiuvyCWXXBIW3k499VR58803xefzyZYtW6wuKhFRXOmGujcSeczMhaGO7CKducLsfu9OMqzcH7sMIpKqRAabsOJ+xWjlSUYmjqlZwTHSeZXI/iRSnnTKHOl8yG/KlyPuPyLosfL55WFzDZrNjher9uzZI/PmzZOLL75YLrroIlm2bJkUFBQEhbkrrriC3SmJyJFSCnUAzgVwJ4B3/ffRaYsXwF+irWfGwlBHdpFu5cxJ9xrF4/T9sXNrYyaPrdNaP80IjtGOgavVFXd/EimPUaOQaufDUQ8cJWXzyoLKWzavTFwLMhvq7HKx6pVXXpGZM2eGtcTV1NTI0qVL5aWXXpJ//OMfGS0TEZEZUg11XwDwIwB/9//Ulv8EcHK09cxYGOrILpzQ3Yiis8sgD2aFtkS3m0hXylRfP/Teu8LmwqDJvVMtsyY0OHat6RL3woOTohsZRl0LXEkPIqOtG6ulrmvNwYnci+4tinl8IslrypPO1Z2Wd78UyfzFnZ07d4rX65VjjjlG6urq5Pvf/75MnTo1KMx985vflPXr15taDiIiK6Tb/XJCvOeYvTDUkV3YsbuR3dmlRS/bh2NPZlAPo1pXIr232mPKq6SwuTBoMJXQ453KwCqho4eWzSuT8vnlQdNCJLsv8SbV1vYnUgBL9p46bSL3w+8/XEoeKJHWVa0ysW2ieFZ40rrPLlOtzJn+LA8PD8uaNWtk9uzZYS1xJ554osyfP182bNggO3bsML0sRERWS7Wlrs3/8xkAS0OXaOuZsTDUkV3YpbuRU9jpeCVTEbZLEE1Govtn1HGI995Gap0qm1cm7jZ34Pla6NPzrPBEHOU0Urnghbjb3EHPSSXcpHufYzKjX2pzE+q7dmqvlc59dsl8rlI5vzP1Wf7www/l7rvvFgDidrvlpJNOkiVLlkheXl4gzM2YMSPleeKc+NkmItKkGurO9P/8QqQl2npmLAx1ZCesFCTOTi2byYzy6MRBYBLdP6OOQ7z3NvR1Ji2eJN0bu8PuNataVBX2msqrEjr2RnWHTndfk6G8SqoWVQm8kKrFVdK15uBci+ncZ5dMoNNaBqsWj02NUNhcKDOfnRlzPSOOQaQy7969W7Zu3So/+MEPwlriAMisWbNk+/btsmvXrqS2G2vf7XCRiYgoFWl1v7TDwlBH5Ex2ugfRjJYsI6Vb4Uz0XjOjjkOyUzPkNeVJ98buoNepWlwlyquCXjPapOaJ7HO858cSKxgYGR4LmwuldVWrVC2qCup6GaulLt0LSfr1i+4tkst9lweda62rWqWwuTDmdo0Y7GVi20T5zarfyMzGg4Oa5Bfky7PPPhs0QuXtt98u7733XlLbTeRzY6eLTPHw4iERRZLuPXXnA3gRwF/9E5FvBrAp3npGLgx1RM5kp0pUopW/VOdtS6fS1bWmK+L9Wskcq0TvNTPqOCTbZdG90C3l88uDXkcbSEV7Tl5TnrgXBnenjBUcMtXyYtR5PGnxpMC9c54VHnEvdMsty24JdEM14x7I0PX19zpqtMdj7U+qx2DXrl2yfft2KfmPkogtcSUXl8jmzZtlz549Ce1POuWy00WmWNiiSETRpBvq1gOYCuAYAEdrS7z1jFwY6oicyW6Vk0SCWCKVxK41XVI6t1TcC92ivErcC91SOrc05cFGJrZNlLymPNn76d6gY5RshTPRe80SPQ5aC5L2PO13fbnjDRCirR9ttNGLOi+SonuLBF6I8iq53Hd5UDniBQcjwnW8bRhxT53IwVAROvhKrJE70w2Uoetrr1u1+GC3155NPVK1qCrmuZboZ3nTpk1y4403BkJbXl6eNDU1CU6AnP4vpwsA+cEPfiDbtm0zJFAlE9TsdJEpFqeUk4gyL91Q91q855i9MNQROVemuxEZ0VUtUuVVP+x/QXOBHH7/4UHPSXWeMH23u1QHzghlRIvEzGdnSkFzgbSuapW9n+6V1lWtUtBcEHTvVbLHOvT5oaNfelZ4pKC5IGjETLMvAiQaVqLtazLrp9Iam+57Gbp+15ouKWgqCLtvUR/Yo4l0DHbs2CFr166VBx98UGpqasJa4mbPni2vv/66VLVXmRJUkglAZl1kMvo7ziktikSUeemGujkA5vknI/9XbYm3npELQ51z8D4AslJoF0RtLjDXAldS52K88AEv5Ni5xwZts3tjt8CLpMusb73RXmPvp3tFeVXUCme8z5kRV/rjtdQZIVI5Exn90khGt4RFWl8/OMnEtonSuqpV3AvdCU1jYEb5LvddHmgZrVpUFVSOeOfW4OCgTJ8+PSi4nXnmmfLAAw/IhRdeKCeeeKI0NzfLzp07g9YzM1Als12j/48yY7/YUkdE0aQb6nojLD3x1jNyYahzBrt1taPco1WG9Odi98ZucS90h7W2hVboYlX2QitZ8ELm9s0NqmSlGur02w4d0CJaoEtmXrRUP4uZaC2wQ4uE0S1hkdZP9j3WdK3pEtcCV6A7bSpz8SXS8qyd75GeW3FvhcxaNEt6e3uluro6rCXuhhtukL6+voTLYsZFPysvJpoRwPh/KRFFw9EvKSN4dTG72a0VNlJ5tAq2/lzUKtihg3IkM3hIaMXd1eqSsvllQV3YyueXi6s1+e6XyVbgEv2cpft+ZeLzHPoaXWu6Aq2r+qBhZhBId667RI5TKsExrNXZP5KpqzW5VmdtW4kcw0mLJ0nH8g654oorwsLb888/LxdddJF89atflXnz5qU8sEm2MevChN2+b4nIHlKdp+5a/8//jLREW8+MhaHOGexw1Z3MYbcrx9HK41owNqGz/lzUKtihw+fr/xavYh4pfHy25bMCLwReSGFzoYy/b3zKxyOZClymPmeZeM9Dg0v5/HIpm1cWaJEqnVsacXAVI7vMxRopNNltRStjKgHZyFAd6/x6++23paurS9asWSMoCR+dsq6uTtRNSg4cOJD06+YCXswkokxKNdTN8P/8UaQl2npmLAx1zsD/3LKX3d7baOVxtbpkYttEcbe5pXtjd1AFO3Sia5GDYShSUOpc3SnwYmzkxtbgkRs9KzyS15QnpXNLx4bhb0t99Euj9t2M9yITrQVBLWYh0xm4F7rF3eYOer4Zg2t0rYk8p1+y+5Dq6JmRmDWx+i9f+qV85rTPhIW3Bx98UA4/+3Cp/069LF68WPbt2yci/A6Px24XvIgou6Ua6h7w//xatOdkamGocwb+55a97NYKG6s8XWu6xNXqCoQErfUldG4ukegtdV1ruqR8frm4F7oD53Lp3FJxLXAF7oWKti2zOWUEv2RFek+VV4WdY5keBt8oyR5fo8L75+7+nNwz/x557rnn5JJLLgkLc1VVVdLf3y/Dw8P8Dk+R1Z8dIsodqYa6AQCFAN6I9pxMLQx1zsH/3LKTU1rqQkccjDWCZax76txtbimbVxb1PiurQ64TRvBLVqT3NLSlLtI9d0a9lt1apFJ5T0ZHR2XTpk3ypS99KSy8XXnllfLwww/LwoUL5dFfPSpqtoq4DX6HExHZV6qhbh6AjwAMA/hYt+wG8HG09cxYGOqIrGWHSn+i5YlVKU30b/BCOld3Br2mPrRFGs1QeZUUNBUEWvMyWSF2wqAo8UR6T/X31EW65y7Vc9Bu53M08d7X0dFR6e7uliVLlkhTU5Pcd999YWHukNJDZPH/WywjIyOB9ewWYImIKDGphroi/8+noz0nUwtDHeUiu10xd0J5wkYMTHGeunghR3sdzwqPuBe6pXVVq5TOLZUj5hxhSOhIhhEBJdmWx2jnQrrnSLT3VBvWP3T0x3TCSarh30qjo6Pyv//7v3LllVeGhbfjjz9e/vznP0t7e7v09vbK8PCwiDgnwBIRUXyphro3/D9/Fe05mVoY6ijXsCKWGi2MAC/ErQAAIABJREFU6Y+ffp46I0c07FrTFTRJtjbypj5oZGIAEyPu70skxGqv51rgijgiZazurYnuU6Twpr0XeU15svfTvUHbdeqon4kaHR2Vp556Su6++265+uqr5bnnngsLc5///OdlxYoVMjo6GnU7dg2pRsjmfSMiCpVqqFsL4HoAfwNwRegSbT0zFoY6yjV26A7nRLHmqUv2+CVSWdS3cGn/1geNTISOvKa8sFEjk33deN1Zg+43XOiW8vnlYfcbFt1bFD6iZIL3v3Wt6ZLSuaVSNq9Mujd2S/fGbimfXx4YnEZ777Tt63834jNhdEhO1ejoqLz11lsyc+bMsPBWXl4uf/zjH+WnP/2pDAwMxAxxZrFbgLJTACciyoRUQ10tgAcB7ADwi5DlkWjrmbEw1FGusXogDqfSKvqR5qnTHz+jKqf6oKH92+yWutDAP2nxJGld1Rr0Oqm8brRjEvp6eU150r2xO2j7Q8NDAi8Cx1yrbHdv7BblVVFbOvVBquSBkqDX6dnUEwiFQ8NDQRX4vZ/uFeVVhs9Zl0pITudcGh4elq6uLrn88svlqquukscff1yOOeaYoDB39tlnS3d3tyUhTs+OAYoXv4go16QU6gJPABriPcfshaGOck22VVYydYVfq3hGmqdOO35GVk5D7+EzaiCPWEIDv1YG5VUpd3uM9d6Evt6kxZOke2N3YPoIbZAY5VWBFq54ATf0PVBeFRicRr9NrStm6KA0WhBMZx+1x+CFuNvcQSE20ZAc7Vya+ezMqMd03bp1Mm/evLCWuIqKCvnpT38qK1askI0bN0YNcVa1ltnxO4kXv4go16QV6sbWRzWArwO4TlsSWc+ohaGOco0dr4qnKt19SbYS27Um8jx12msaXTkNut+s1WXo6JeR9j1S+T0rPEH39xl572Do63Wt6ZKyeWVSOrc0MEjMxLaJcrnvciloLghMzN69sTtoW9FGD9V+P3besVI6tzToXsiyeWWS35Qf8R6+dPZRP6qm8qqgsiYTkqO9F4XNhWMtivv2yqw5s6Tg6AI584Iz5YYbbpAbb7wxKMxdcMEF0tvba9j7ZRblVVK1uCroHLMyQIXe02rE4DlERHaXbkvdjwD0AviHv+vlBwCeiLeekQtDHeUiu92/kqp0QlQ6ldhox88JE09r60RrBTKqYp/IexMtFBU0FYTtjxYu4YUUNheK8qrA3/XbjdTaWPJAicALmds3N+ieOs8Kj7haXSl/FiLto7vNLe6F7qC/68uXaEgO3Y/R0VE5qfkkURepsJY4AHLnnXfKpk2b5J133km4/PH2xewQo79Icuy8Y4MuknhWeCwJUNo5qR99VvvdqRe/iIgSkW6oGwCQB+Av/t+PBfBMvPWMXBjqiJwrnRBlRiU2ExXjREaLDK14hgY/V6sraouc8qrAz3QCf6LvTaRQGm1d5VVSOrdUyueXBwY90Vr2ot2np+2b1t1SP/pluoE7UjnzmvJEeVVg37T7/7QBdbQBWuKFusq2Smn0NkpJSYlMmDBBjj32WMHXg4PcV77yFen7Y58hFw0yfUFC3515Wtc0KWguEHghx7ceL9N/P10KmwstCVD680ffjTbS+UNElE3SDXV/8v98HcBhABSAt+KtZ+TCUEfkXOmEKDMqsWZ3YUt0tMh4rWGhE6Cne+9cJOm8N9HW1UbB1AdBd5tbXK2umPs7sW1ixCBrRojXt9Rp5XEvdAfCQLQQPjQ0JDt27JA5c+ZEbIm7tP5SKbizQG77zW2G7kOsfYm0baMHAtIGpfGs8MiEhRMCLbHwIuV9SUeklt7SuaUCL8JGT2WwI6Jskm6oWwLgCADfAbABwJsAfhFvPSMXhjoi50onRJnVqmZk19awFrYFwcEk2miR+mAaMXgsdIu7zR30HCNGuQwtezrdWyOtq4VOvURb/4wK3KH3OYaGNP09daGvo38vdu/eLXPnzpXizxYLABk3bpw88cQTUlhUGAhyn/n8Z+SYO48J7IeR3WMj7Vci8yca9fpaeCq6t0haV7WKyMH3snVVqxTdW5T2PqUi2j2Z+s9Lz6ax0VOz+f66bOmiT0SJS3uglMCTATeAzyezjhELQx2Rs6Va+bDzgDHRBmQJbWHTjxapCQ1jkVokO1d3BkZ+NGo+umj7kWrFMNGBXEJHvoz1eulWVKPdAxjanTLS6+zfv1/UXUpaF7TK5z73ubCWuJkzZ8r9v7tfJtw/IW6wMrKyrd/eUQ8cJQVNY90gi+4tkpnPzgx6bqoXQmK9l8qrxL3QHRjAxr1wrKVTeZUlwSL0Pc5rygt0C9XK0bm6U/Ka8rJ2JEw7fzcSkXkMC3VWLQx1ZAe8KmoNfXiCF+JaYP29Mvp7jUKnTojUta9sXpm429wJjzApMlYRdy04OEBIpifFNiOIZ6Iimkyo2blzp3i9XsnLyxMAUlxcLEeef6Scf9H5UlNTIwDk2muvle/88juBgVOseB/002aUzSuT8vnlUafNSKXLcrxBedxtbrnnxXsC8wa6FrgCo51qU0KYOY1HtDJr52dBc4HkecdaD7Xya/dxJjogk9O+260YNIeIrMdQR5QmXhW1jh2PvVahijTJeWgLW6ClqDX6wBtmdqszcuTNZIJdpNfMREU0Vqj55JNPZO3atXLHwjvksyd9Nvy+uMmQkltKpOT+ksC+e1Z4AlM1mNliGo3+mEUaqTP0+KVyjGOtE6lFWjsmJQ+URJ0PMpNcrS45Ys4RUjavLHBP3VEPHCX5TfkJfT7s9v2SCM7RR5SbGOqI0sSrotax47HXKlT6smkVqp5NPSkNwZ9I+Eo2oKVaYTWyC59eJiqi+rJ/+OGHctdddwUFt7zxeTL+6+Pl1JpTxeVyyRmXnCH5d+QHQltoCA9tmTPj3sZY9MdM+7f+mIUev1Te80TeF/17qx2TSBc1rAgWeU150rm6U1wLDrboH996fGCE01js+P2SCKeWm4jSk+5AKUdFWArjrWfkwlBHZkm0kpztV0W142DUUPlGsmIY93jnhFah0leg73nxnsCIgFYMp57K/W3RGNGFT5uEW3lV0H1ghc2FclHnRQdHxlzoFtcCV9TtJmP37t1y94N3y3GzjpNJZ04Ka4n793//dznmxmOkbG5ZUOthrJAWaaRFo0chjSXZljqtjMmE/2TPk3gXNTIdLNIJOHb9bk/k3lMntjASUXrSDXVvAxgBsB3ADv+/twJ4A8CZ8dY3YmGoIzMk859iNl8V1Y6DNnFv66pWcS9022Yi30we+0TPidD7nLTh1MffNz4wWENhc2HYIBZmiRao4IVULapKuqtgul34tPK0rmqVI+ccGRjgY8bSGTKuZZzAC5nWNS3toefff/99aWxsDAtvsxbNkkMmHCI4BYJzIEd6jgxsP3Q00rymPNn76d6gY6LNtzdp8aTAPWP68iU6ObkRkr2nLt3XSCQgRLqo0b2xW9xtbku+M9IJOHb8bk/me8hp9wISUXrSDXUPAbhY9/tFABYAOAfAa/HWN2JhqCMzJPOfeTZfFdWOg/546H+3Orhm8tgne07ou6Nd7rs8qJytq1ozNjlztEClDUWvP16JdqNM9Jhrx0EfIF0LXIHREeGFTP/99MDrulpdcuScIwVeBCqiiZ5nO3bskGXLlsm6devkvPPOCwtzFRUV8sILL8gvXvuFlM4tDQzsMWHhhEBwnLQ4eDTSSC11Wig2M0glQ3+uuRa4Yt6facRrxNtuaNDUz/Fn1XeiGQP7WMWOQZOI7CHdUBe2svYYgNXx1jdiYagjMyTb7SZbr4pqx0F/PPS/G9UNKZ3jl6ljn2pXrLymPKlaXBVUEdO39phNX26tQqi9vhbwqhZVJVVhTfQeP61CXLW4SlpXtYZNAq0fYEMbYn7XJ7uCJq6Odv+WukvJEecdERbePB6P3HLLLXLnnXfKPffcIx9++GFQuVytLimfXx5UrvL55YHQoR+NNHQglJ5NPVLYXBh0H13XmoOTk9v5s5/J76hs+j60277YtUsoEVkvVqgrQHz/VErdDeBx/+/fALBTKZUPYDSB9YlsqbKkEn1b+lA3sS7wWN+WPlSWVEZ8fn1NPepr6jNVvIzRjoP+eOh/j3Y8kuEb8OH252/HuEPGAQD2HtiL25+/HQASOqbJHnvfgA8tK1swuH0QlSWV8EzxJLR+rHMi1jYrSyoxuH0QtRW1YesNbh9MuNyp7oe+3Fo5+rb0oaq0Cp4pHvx45Y+xbvs6NC5vRMuFLYYd85aVLeiY1oG6iXX4ryn/BU+PB3kqD4V5hRj4cACL/rQIBaoAU0+eitm9s1FZUomPPv0IE//vRABA9ZJqeKZ4UHZoGSpLKrF582a0L23Hr//ya5w7fC7e+tVb2IVdgdc75JBD8MILL+Dss8/GZz7zmajl2rp7K7qv7Q4q1x3n3IG7XroLZYeWYURGMDwyjOKWYlSWVOI7Z34HT61/Cvf33Y/KkkoMjw7jR1/4UdCxuKrqKhS3FGPtrWvjHrtkpHquRtqOp8eDjmkdgfe/YWlDoPwUnd2+25P9v4mICEBCLXUlANoBvAlgNYBFAEoBHALgpHjrG7GwpY7MYMduN1bIxD11+pYT7VhrLSehZUn3ink672u0dbX5uqJts2tNlxQ2FwbNk6Ud01Rb6pLtAhloMVtUlVKXy2TLFtrlUntcG31Qa+265ZlbJM87Njl06D11v/vT7+QzZ38mrCXulOpT5PHHH5empiZZtGiRLF29NKnywwv5wYs/ONhlsdUlJXNLEm5py1T3NyO/g+x47yklJvR7L973DRHlLnBKA6LI7NbtxmzR9tfs0S+17nh6Wtc8fdmMqMikW7lNdRTJmc/ODIz2WLWoKqlQbMTIlfr3UAtUZlQII3W51G/fvdAdGFhE26fx940fC3vXQfIuzpOJl06U4791fFiYKy4uFlwNOa3ttKBjkWzXM22OstZVrbL3073SuqpV8pvyZVzLuIQ+72aEFiPe41gy2WWP93wZJ9aFpFz6v4mIEpNWqANwCoCfAegG0KMt8dYzcmGoI0qflVfXtVCnr9hOWDAhKNQZVVE0o3Kb6DZTuUgQ7X3RhsxPZT/MvFgRaVCW6b+fLoc0HyLKqyS/KV+K7y0Wd5tb1HeU5B+RHxbexo0bJ6+++qrcd9998sT/e0KURwW2Z0SLp2uBS/5/e3cfH1dd5/3//U0aWkFuxERbCGXiXqumSb2tq7vbXmu6utpLrbiy60ZkcY3rJWwj6qXV3XEhLUaxkLYYWvdCg5ILE/2J4lXFYrlIcFu7F1zFLU2luGArXRSVKoKCvc3n90fmDDPp3M85c86ZeT0fj/NoM5kz851zTpLv+3zvzvz0melxcN5MoBpQyWHXz2MYxDmerZZBK8wxX/V2I46ADKAc1Ya6+yRdKumPJL3S24rt5+dGqAOqF2blwatkz79mfnoCjbOvnmlNyZxq3o+KYhCfM8hjl++15141N9DzVWnlePZ58looNSA77/3n2bl/fK5piez0S04/KcxJsov+4SI7evSZ/TM/U9emrnQLZzWziM5ejDqxMWHnXHtOeuKYoLql5lPtOS530poozRLrp3rs9smkKADKUW2ou7fYc4LeCHVoBKVW3Cq9Sx323fXmNc32/Gueb27AWWJDwtrWtWW1wPhVUaxkOv4wuuN58p0XL4AE8Z7FPk+h45J5nnbu3Glzz517cjfK0+bZee87z171qlfZhg0b7Lap29KvU+h9vWOR+f6Lrl9kbsCV9fm8MmaW1Q249Ayl3jVVq+u/mnMcxPVcrbDCVT22atXjZwIQnGpD3YCkyyQtkHS2txXbz8+NUId6V0olqdqKVNiVB69SnW+slJ8VxSBaNnK9ph+V6ELnJahKerH3zHdcbr75Zpt72lxzL3emN53cCifNLPb95NNPnrSYd+Z5LiUwzi5XObzP4AacPXXkqaxlCjLLEnZLXbFzPLZnLOci51Go9IfRDTLIbtBhqcfWRwDBqTbUHcix7S+2n58boQ71rpTKrB8TgIRZeSil/LWsjEXleIZxXgpVjtuH2i2xMWFNa5qs46Md1rmk8+Tg5mSt/63V3Oud6a2ylf9rZfp1ujadvJh3qcfV72CfGYi8GQWHdg6lW+xqdf1X8rkyg+mNP7gxa4Fvb82/RlPq75C4haQ4hVAA4WL2SyDiSrkD7Uf3yTArD1GrbFV7PP1s+axlt7muTV3pBcEz3+fOH99pZ/3lzELf3a/qtresfIud9uzTTgp073jfO7LGxCXvTFrL2pb0ec21mHc557mcLrHltsZ6ZfV7dtdSlHuOveurfX27zb9mvk3sn7BtD22zxMaEzb9mvrWvby+4f7XvH0Wl/A4Ju0cCAASpolAnaXnq37/MteXbL4iNUId6V4uWuiiIUsWy1OOZr8xxm+Ags0I8unvU5l8z31pXt9oLul6QsxvlO97xDrvnnnvsnz/7z3b++vPTxyDf2LDMY7TqtlXpSUq81iU/z3UUx5n5zTvWmWs8PnXkKWta05RzjcdConZDpRrFzmfcfi4BoByVhro1qX+/mGO7Md9+QWyEOtS7WoypQ7Zqj3lmKMysaM69am4kz8mi4UX296v/Ph3anvO859jZbzv7pDD39n94uyXWJ9Kf2VtPcPZn9uQLwn5eq7Mr8u1D7bG/wVGMd6y9mTy7Nj2zjmS53S/jdkOomiAet88KAOWg+yUQA0HPfllLQU0qUotyZiplQhFvCv6hnUOW2JAoa9HxoB07dsy6u7tztsT19vba3vv3mnuns+npaTN75vNmHpfEhkS6q1+pYc3vrqmz31MDstHdo1nPq7fWGO9zt36m1c7fcH56UflVt60q+1jGqfWq2hsC3PwCUM8qban7cKEt335BbIQ6ID5yVara1rVZ27q22FW0ilWGvck4Zo/RCqNl4Pjx4zY2NmYvetGL7JxzzjEzs+np6ZPC3CX/eIktun7RSYHN+zyltF4WC+d+hohcATGxIWGJjYmsx/w85n7cgPDjNVbdtsqa1zSbBmTnbzjf/vGOf7QF1y6wtnVtddt65UdZo3gDKU44fkB0VRrqrkxtY5IelDSU2v5D0hfy7RfERqgD4iFfyElsTFhiQ3CV8KDk6mLpdYGrxdi6UipXb3/723O2xHmtbw8++KBNT0/b2J4xa1vXlrUAfK6A4EeFzs8QMfv4ju0Zs3OHzk0vKj66e9TXmwR+tPT41VpUrOW01uWphWp+nggj1YvTtQI0omqXNNgm6fSMr0+XdHux/fzcCHWI2x/ruJXXD15loGlNU3ptMK8y0LSm6aQFpKPa/StTZhfLxIaEDe0cyupyWc44s0rfOz227Ufb7NkXPtuaz2o2SfbioRfb2J6xrCD32te+1iYnJ/O+ZuZyBd51GUS49rNiODtYe11dzx06N2uaf79+xvw4n35dE37eMJj9O2nVbasi+Tuq0mNHGPFHnFp1gUZUbah7QNLcjK/nSnqg2H5+boS6xha3P9ZxK69fvMpAZqXA+zquLXVmxReADup8e8dxaGgoZ0vctf/rWuvY2GFfvPuLJb9mLcdWje0Z82UGzMzju+j6RelgHVRXVz+OUTmzhHpftw+1W/v69qzrrFY3DKL0O6rSshFG/BGn8ZdAI6o21CUl3SdpINUdc7ekfyq2n58boa6xxe2PddzK6xevMpBZKXvqyFPmBtxJY+rCXC+sEqWMrau21eP3v/+9DQ8PW2tr60xw+zPZ0eNHbWBgIB3k/ui1f2TuvS5vV9Bianlt+hkcvM+rAdmiTYuyXsPPCmexAF+qXMe50Hp+o7tHbcG1C2z+NfOzupN6C6YXO4blXn9R/x1Vyc8TYcQfUb82gEZX9eyXkl4p6fLU9vJS9vFzI9Q1trj9sY5bef1SbIr/zCDSsral4sWpw+BXRSdXZXXfvn05W+LOfPWZNrF/Ij02rmlNk217aJu1r2/PCs1Na5pKPn61bKEJonJY6mtWEgpK6WpbqlzH2bvmMz/L0M4h69rUlf5cmZ8lsyW40Gep5JzW4+8owog/otyKC8CfUNcs6RxJC72tlP382gh1jS1uf6zjVl6/1Gq6+zDGK/pR0bnx/95oz1n5HJt36rx0cOvY2GGj946mv77wwgtt9+7dOd8zsSGRXnR6dvfWUgOA97qVHr9y9i21C6LfE5D40X2v0pbQ2WXN/JxuwGUdD2/sadOapvSxygxWpYasSn6e6vF3FGHEP404JhyIi2q7X/ZLOiTph5L2SJqStKfYfn5uhLrGFrc/1nErr59KqQxUO7tdWMe23IrOsWPH0v/P1RJ31llnFa1IZ75n+/p2a1vXZm7AnTQRjReWgjw25R77UrogVlLGYueh0sASdOvV7JksNSB7/jXPt/ah9oItdcVUUu56/R1FGAFQ76oNdQ9Jem6x5wW5EeoaT1xmasuHykV+1bQS1KKFodJz9/jjj9vAwIA5505aVsD7+qJ3XWQPPPBAep9yQ0OhMV9zr5ob6LEpdOxzHbNSuiDOLqMfPzeVhrOgr62xPTNLSiy4doF9/I6PW+tnWk0DsqaBJnvdTa/LOaaulM9fabn5HQUA8VNtqJuUNKfY84LcCHWNpV7vImNGNec36NaUcsr2q1/9Kt0ad9555+VsjfvNb36Tfr6f4/JylXF29z4zf49Noe6U+Y5ZsS6IZmaju0dNAyprrGWhQFJNyAn69077+nZrXdeaXmPv0m9favOvmW8akJ39mbNPmv2yFPy+BIDGUW2oG5G0Q9I/SvqwtxXbz8+NUNdY6nG8RyMqVPGutJUg6Guj0Ov/8pe/tNWrV2eFts9//vNmZvbxj3/cJNn73vc+O3DgQM7XDmI2yFpMf++Z/fpje8bS68QlNiSyPsfE/glrX99etIxje8ZswbULLLEhYYs2lbZcQbHjWM1xDrr1qmlNky26flHWMfCCd6ljIsMoNwAgGqoNdVfm2ort5+dGqGss9TgzW6MJqvUg6FaJzGvv0UcftYceemimNeqNLmdL3B133FF2+TMn4PBzSYdSj001wcF7/dHdozb/mvm24NoF5gacbXtoW9Z7ea1vs8sye4r+xMaEzb9mvo3tGUsf+8wgl+vnvpTwGtWQk6u1cmL/hC3atCjwMZGoXFSvJwCNp+rZL8PeCHWNhZa68PhVeQnyHAZVwXrkkUfs7NeefVJwm9g/YS/81AtNkn3wgx+0n/70p1WVM8hgWuzYVPvemZN8JDYmslrfMs9vYkPCEhsTWfvmGn+nAdm5Q+eml77wul56QS7XNRPXmz5je55ZjH322LnkncnAx0TmKg9BpTi6twKIkopCnaSNqX+/JWnL7C3ffkFshLrGwh/RcPh53ONQ8X744YftlltuMTOzI0eO5GyJu3zN5RUdg0LHsn2o3RIbE1mV6VrdtPArbGeeX++zbntom7kBZxP7J0wDstHdo1n7zD7/Y3vGrGVtiw3tHMpajPuir1+U7qIYxJIYYci8Hi791qXWvKY5HWi9dfCCHhOZrzxB/o4NKjjWMpDG8XoDUL8qDXWvTP37Z7m2fPsV2yS9SNLujO1JSR8stA+hrvFwF7n2/Ky8RLEitH//fnvve997UnAzm7ne5v2XeXbRZRfZpslNltg4M1asfX17Rddevs/vtdRse2hbVmV6dPdo1ZX3Un5mmtY02eju0aznVfLeOcfXpY5Z16aurLX0PLPPf9emrnSg8cLFRV+/KD1pSr7PEMebPoXGI5Y6JrKc34nFnlurWWTj2AV7tjjcoALQOPxYfPwUSS+RtFjSKaXsU+LrNkv6uaTzCz2PUAcEr9L1rmrdxbBUDz74oH3gAx84aVmBzG3NmjVm5n8lN9+x9CYVyXyvif0TltiYqKpCXerxbh9qtwXXLsh6nregeanv440LKzRTZSnl8Y5R5jXkjS0rtRxxuelTys9WrmPWtq7N2te3lz0zaKnHvlB5qhVUcKz1DaMo3qAC0LiqnSjlTZL+U9Jdkr4n6aCkFcX2K2WT9BeSvl/seYQ6IHhey0lmZdn7OpdSZiGsZcX7oYcesne+850nBbfvf//7Zma2adMmW7t2rT3++OMn7etnoDXLXxHUgOzca89Nh7vR3aO27aFtpgFVdXxKrXi2r2+3+dfMzzpn86+Zb+3ri4e6Vbetspa1LeYGnC26fpFdMH5B+ut84/fCbi0qtSxBK/WzZpazfWhmsfmJ/RO26PrSZgYt9b1qceyDCo61bjmLwg0qAPBUG+oekPRfMr7+A0kPFNuvlE3SjZJW5fne+yTtkrRr4cKFQR4fADZTaZ+zdo4N7Ryyp448ZUM7h2zO2jm26rZVOZ9fbsWwkop1oX2mpqbsr//6r23nzp1mZvYXf/EXWWHulHmn2PPe+jxzyeIzTFbyWcqdVr9tXZs1r2m2oZ1DNrp7NN1dsXVdazpUVRo+ClV0Z09Mcum3Li27++Xs8W+ZE3xUGgRqOa4r7Er52J6Zhce9sZSJjQlrW9dW8jVZ6sygmc/NVEqroN/HpF5a6szCvykAAJ5qQ92/zvrazX6ski3VpfOQpOcXey4tdUBlyqmMlNtSV84d80oqkbP3uXXXrXbqS049qSVu7oK5NrZnzB588EG76H9cZJ3rO8vqrlZJ+UqpWM4+9u1D7SeNIRvaOWQta1tsbM9YVRXtQmP48i0hkK/c+V5/9vn2WpCqaSGpxUQa3qyamWrdfS4d6jYkzA04S2woHuoyj7d3fovNDJr53EzFWgWDCCrVXM+FyhaFkA4AYal0opS/TG2fk/QdSe+WdImkb0sayrdfqZukt0raVspzCXVA+cqt/JTbramcO+aV3F1/weoX2GuWv8Yuv/xyMzP78Y9/nBXmzjjrDFu/Yb1tvX/rSWuglbOQdbq72/p2ax9qL6mSW0kXsGJjyKppgch3rs+++uysmTZX3bYqvdh3ORVir6yZ5Tt6/Gi662UpatHa4c0umtm91Vv/LfP9Sumu52d5Kzm3mft453dknjfWAAAgAElEQVRo51D6PFQzpq5WKm2dL1Z+Ws4ANKpKQ90XC2w35tuv1E3SVyT9XSnPJdQB5Su3Ilnu870xVl6F32uFyjfrYikh6MILL8w5oYln9OZRU1I5y5m5zlcp3dWCaBkrtZKea59ix6hYRXb291fdtirnTJuXfvtS04DKqhB7rbb5WhmLqUXQ8N6jbV2bnb/hfGta02Qta1vs7M+cbUM7h7LOTbFz5Xd5Kx2zmVmG5J3JgmMYZ+8b19DDxCQAkF/Vs1/6vUk6VdKvJJ1ZyvMJdUD5yq1IllOR9Z6bvDNpi65flO7uWM74u41f2Winvfg0mzt3bvqxzCA356w59uFPfdiOHDmS/r432Ui+mSXL6a5WSeWx1Nkf8+1b6PgWKk8lIaNrU5dvM21mne9Ni9KBKd/5zlWWUo91pYHEC56ZQXZo55A1r2m21s+0ZgX9Uo6dn8Gi0teLYzirtswsIQAA+UUu1JW7EeqA8lUTWopVyCqdWOSiyy7K2RL3u9/9zszM9u7da0ePZi9qPTvItK/PvQZaZktdVne1PAtZVxt6y2k5yXyNSsYKta9vt8SG8hYs99aky3zNambarKayXuqxrqaFzGsxTmx8Jsh6XURb17Xa3Kvmllx2v4NFlLpEBqmUCYSKXUO01MFvcbw5AuRDqAMiLMjJIoKqSBar9E5PT9t3v/td+5M/+ROTZFNTUza2Z8zmnDUnHeTmnz/fbr75Zjt27FjBzzD72OT7XJlj6koJXUF3T61Evs9byYLlXnkzXzOxIVHS8gV+K/XYVXOMvc+YGWS3PbTNzl9/ftlBNohz3QgVSz9amxslAKM2uJ5Qbwh1QEQF/QfH74qk93reRBSzZ1E877+fl7MlbmhoyMzMnnjiCTt+/HhVZcgsR65Fz0t9POiJZPzStakrq/XJLNWNckPhbpSFPl+tA0apx7qaYzy255llFzKXjGhb11by4urllrdW4hIIC52/WnTBBWaj5Tf6+HkvT6UTpXy40JZvvyA2Qh3qVZz+4GRWdG/695vsOe9+js05Z6blbdkbllnHxg772OaPpYPcS17yEvv6179uJ06ciEy5qwk3fp6rct63mm6UfoRZv5Qy0UuuLpLlHOPZC6QXmryn2vLWStQCZiGFfkYYK4dK1aLrN8IRp99vUVFpqLsytY1JelDSUGr7D0lfyLdfEBuhDvUqTn9wujZ12db7t+ZsiZt7zsxacdPT0zY9PR12UbP4Fcb8+uOT63Xa1rVZ+/rcyynk7Ea5MVF265OnkvF51Sol0HkTsSQ2JNLLUVQSyqISxvwS1xs/s39G4vQ5EB3V/t7luos2zk/5ql18fJuk0zO+Pl3S7cX283Mj1KFeRfUX2vHjx21sbMxe9KIXpYObF0C9r1/96lfbt779rfQ6a1Exu1LvBpxvwTnztctZ1y7T7HM+tmcs79px3iQpGpAlNs6su1bNncxKx+dVo5RKWeYx8Y6xG3A296q5kQpllQbGRmppKNQFmjvyKFe1fyO57qItbr/foqDaUPeApLkZX8+V9ECx/fzcCHWoV1H5g5PZuparJU5SzjDijVuKSotIruPpLT2QyY9JLyo5b16oyqzwdm3qsm0Pbcv6Izaxf8Lah9rT7zG6e9QSG2aOdftQe8XHutLxedUopVIWhz/s1ZxzWhpm1FsrKoLnx+8Grrvoqqffb7VSbahLSrpP0kCqO+ZuSf9UbD8/N0Id6tHsFolypsav1tGjR+0LX/iCLVy4MB3afv3rX5uZWXt7u0my5cuX21133ZVV3syQseDaBTb/mvlVtx75KdcfCG8WTD+DcyV/iLzjl9iQsG0PbUuXww042/bQtqx9vXX3ynmPUioulY7PC7qlKayZRctRaRlpaQAqR6W/vvH7rXxVz34p6RWSLk9tLy9lHz83Qh3qTa1/kR0+fNiefvppMzO78MILc7bE7d27t6RyF5r9Muw/tPkChBeY/bpTW8kad15wb1/fbm3r2tJT7resbbH518w/6VjmW2Q913uUej1VMj6vFi1NQf88+PH6lbYY+NXS4HXDrba1FvFA69IMKv31j2u9PH6EuqWS/i71/zZJHaXs59dGqEO9Cfru4+9//3v77Gc/a62trenQ9t73vtfMzG688UaTZG9605vs7rvvruj1g1ic2Y9f6kEc11xlK3d69o6NHda0psmeOvJUemKUs68+O11Jb17TbMk7kyUtsp7rPUotTyUVpFq1NAX5h92P6yKsljozKraNhvOdjUo/8Ixqu19eKelbkv4j9fU5kr5fbD8/N0Id6o3foeipp56yBx980MzMtm/fnrMl7gtf+ELV5fZU2v0w6AkU/K4M5Xu92QudF3of71hlHjOvS+jQzqH01PuzF0ovZ3r+cq6ncitI9TCmxa/PUO2YukrHRtIFrbFwvgHkU22o2y3JSfr3jMf2FNvPz41Qh3pT7R/t3/72t/aZz3zGTj311KzgZjYzc6Uku/DCC2337t2+l92s/ApuLac69zNAFCpbqe/jBYrMY9B5fadpQFnHLHlnMr1Om9dFM3ln0hZtWmRNa5qsZW2LrbptVdnlrFY9VDD9+gzVzH5ZzSymcZhIBv7hfAPIp9pQd0/q3x+k/j2NUAdUp9xQ9MQTT9itt96a/jpXS9xHPvKRWhXfzMqr4BaqVEe5AuNH2XJN168B2SlrT8lqrfTWjjt6/KglNiRswbULSh6zGGR3rbC6gvkZzqPQna2aYFkPwRql43wDyKfaUPcRSf9T0n5Jfy/p3yR9oNh+fm6EOtSjQpXWxx9/3AYGBsw5lxXcjh8/bmZmb3nLW+ziiy+2Bx54IKzil6VQOAqqAuNHKAhqPNTsZRa6NnXZ0M6h9Os2rWnKOSNmoTAZZBfHWs/UGkQIi3MX0CiEUtQO5xtAPn5MlPJ6SddIulbS60vZx8+NUId6d+jQIUsmk/a73/3OzHK3xL3nPe+xI0eOhFzSZ/jVUhdUBd6PyTn8Ktvs95k9Js8NuKzX7dqUe+26MO/U17KiWY8tFX5MOMNkEY2D8w0gl2pb6j5TymNBboQ61JtDhw7Z6tWrTwpuGzZsMDOzsbExe//7328/+clPQi5pbn6OqfO+72cFxs9p9IOqXGW+7tyr5ma13I3tGbP518y3xMZEZO7Utw+1W2JjIus4BBW0otwlt1K0vgAAqlVtqPtBjscYUweU4dFHH7UPf/jDtmXLFjMzu+KKK04KdP/wD/9gv/jFL0IuaWn8nP0yCFFZ8LpUuSr8bevarH2oPRJ36sf2jJkGZNse2pYVSEZ3j/q6jEXmmMMoroNYLVpfAADVKBTq5igP59ylki6T9AfOuT0Z3zpd0s58+wGQjhw5oo9+9KMaHh7Oenz9+vUyM33oQx/SE088odWrV+ucc84JqZTlGZ8a1+D2Qe07tE/TNq1Hnnwk6/tLFy7VvkP78u7fu7hXvYt7gy6mJKmztVM7Du5QT0dP+rEdB3eos7Uz/fW+Q/u0dOHSrP2KfYageMelf2u/9h3ap87WTl33xutqdryKGdw+qMSZCc1pmqOW5hb1dPRoZOWI3rPlPVnHtFTjU+NKTiQ1snJESxcu1Y6DO/SOW94hSfrqhV/VI08+oo/9n4/pw9/9sI5PH1f7Ge3q29KnweWDfn+0mqrlzwBQqczf9Z2tnUouS3LdAnGQL+1JOlNSQtK4pPMztrPz7RPURksdou7hhx+2Sy+91C6++GIzMzty5MhJLXGrV6+OTUvcbLNbkhIbEzb/mvmRbEkpdfr4KLXU1VIlrUVNa5psdPdo1jWw7aFtpgH5NgFNYkPCEhsTWeVMbJxZ041WLaA26CYMRJuq7H75GkmnZ3x9uqRXF9vPz41Qhyj6p3/6p5wTmnhuuukme+yxx3x7vzC7bs2uhI/tGbMF1y6wxIbSx3zVovzlLPTciJWXSj+zd/4zz2FiQ8La17dXVI5c3WPdgDupK2fcx9EBcdOoN7uAuKg21P27JJfxdZNyjLMLciPUIWw/+tGP7G//9m9Nkk1PT5uZnRTmrrjiCvvVr34VyPuHHUByVcJHd4+aBlRSSKtV+cutkDTaGKdKK2x+n7/29e3pdfm84z67pa7UsgHwTz1OUgTUk2pD3e4cjzFRCureyMhIzpa4e++918zM9uzZY7/5zW9qUpaw755W+/61Kj8VksKqXSvNjwA8tmfM2ta12YJrF9i2h7bZtoe22fxr5tsZnz7D2ta1+RIcGy2sA34J+28NgMIKhbqmEobd7XfOfcA515LaLtfMQuRA3di7d6/e8Y53yDmnW265RZL0ne98J/39efPm6dOf/rSefPJJveIVr5AkLV68WGeeeWZNyhf2pB7JZUn1benT5IFJHTtxTJMHJtW3pU/JZcmS9q9V+b0JUjLNniClkVVzfHoX92rvZXt14ooT2nvZ3oonThjcPqivXvhVDf3FkD703Q/pjV9+o+a1zNMZp5yh6954nfq39mve4Dz1b+3X4PLBst/Hm4RleMWwDicPa3jFsJITSY1PjVdU3jCMT42re3O3mtc2q3tzd6zKjnir9nc9gBDlS3veJul5kr4i6ZeSfiFpTNLziu3n50ZLHYJw33335WyJ6+3tNTOzJ598Mr0YeNiicPe0mtaPWpU/7G6qUReF4xN0a2oUflaqEYVzhMZGSzcQXaqm+2UUNkIdqnXPPffYm9/8ZpNkL33pS83M7Kc//Wk6yD33uc+1z372s/b000+HXNLc4l7Rq2X5azUhS1wrPWGXPejQFdUuuKUe97iHUgBAcCoKdZJWp/4dlvTZ2Vu+/YLYCHWoVK6WOGXMUHns2LEQS1eesCvj1Yp7+T1hBOygjl0YAXjVbasCPX5RDEXlXDNRDaUAgPBVGurekvr3klxbvv2C2Ah1KOZ73/ueLV++/KTQ5n3d3t5uN9xwgx05ciTEUqIe1Do0BBEiS13LL6iyr7ptVWBh0u/jlS/4lhOIy7lmwgql9XLTBQDqGd0vUVemp6fTywp0dXXlbIk7evRokVdBPctVQfWr0lrrlhS/K/le6ElsSNi2h7ZlhR6/w0PcA0qhUFpOcCznmgmrJTjO3bsBoFFU2lL3LUlb8m359gtiI9Q1tunpabv99tvtj//4j9Oh7d/+7d/MzKynp8ck2R/+4R/azTffHKvulPDf2J4xax9KtUBteKYFqm1dm2/T5ddy0peuTV2mAdmiTYtOCqYaUEnlnx1w2te328T+iayg4ZXf73Aa966E+c713KvmlnUNRH39xCh2WQUAnKzSUPdnqe06SV+V9JbUNibpU/n2C2Ij1DWW6enp9IQln/rUp3K2xH37298OuZT50Y0pHOkWqI0nt0D5ubB1LVo1Mt9j0fWLbGjnUFYw3fbQNktsSBR931xl1YBsdPdoVkXeW0jeDTibe9Xcuh7fVo58oVQDKiusRr0lLO7hGwAaRVXdLyX9aymPBbkR6urb9PS03Xrrrfayl70sHdq6umYqfTt37kzPWPmNb3zDTpw4EXJpC4t65a2eeQEiVwuUG3AnVVCrqbQGHdwzw5B3Tc2/Zr6dv+H8srpL5gpViY0JS2xIpF83eWfSnn/N823+NfOtbV2btX6m1TQgax9qr/pzxf3nYfZ58M65G3CWvDOZ9dxi5yLKN3vyfU4/Az4AoHrVhrp9kl6Q8XWHpH3F9vNzI9TVlxMnTtiPf/xjMzP79a9/nbMl7uMf/3jIpaxM3FsmoqqUCrEX5jLPgRfc/Gypq4XZLSdje8ZMAzINKOvzFwumuVpgvFa5if0TNrp71FrWtpgGZGd/5uzslsCNxVsCSxHlMFNMZvDt2NhhQzuHLLEhYReMX2Bz1s6x5J3JWIbV2fJ9Tu/ruH4uAKg31Ya6N0o6KOmu1PYTSW8otp+fG6Eu3o4fP25f/vKX7YUvfGHOZQUk2Wte8xrbunVregKUuKIbk/9Kbe3xwlzm871uin6OqauFQi1smSppqZvYP2HtQ+1ZY/Nmd8f0rtkoB99aGdszZnOvmmtuwGWF0uSdSZt71dyCYTVOgTbf5+QaAIDoqHr2S0lzJb00tc0tZR8/N0JdvBw7dsxuu+229Ne5WuLe8pa3hFjC4NBS579Sj2lmmBvdPWqJjYmZboTr26ua/TKMinmuIFtJMC0lEBfqtuqFu0qOQZwCTTGV3KzJbP1atGmRNa1pspa1LbbqtlVBF7di3JQCgGirtqXuVEmfkPT51Nd/KOnNxfbzcyPURduRI0fs85//vLW3t2cFt0cffdTMzC655BLr6emxu+66K+SSBi/uY4iiqNzp4P0MEmGeT7+WZSi2T7ElDtrXt5d9DOrt56CSmzVdm7rS3Re94zC0c8ha1rZE9jhwUwoAoq3aUPdVSasl7U19/SxJu4vt5+dGqIuWw4cP2+c+9zk7ePCgmZmdf/75J7XEveENb7BDhw6FXNJweJVobybB2d2ZUJ5aLyGQGX7qvZKbea3OWTsn52Lk7UPtFQWaejpuY3vGrG1dmyU2JMwNuHSX3kI/001rmmzRpkVZx+Ho8aPp3wdRVG9hHADqTaFQ16Ti/sDM1kk6ppmBUL+X5ErYD3Xi8OHDuv7669XW1ibnnObNm6dLL71Uq1evliSNjIzoTW96k+6+++70hXX77bfruc99bsglD0fv4l4llyWVOCuhrRdt1ZFPHNHwimElJ5IanxoPu3ixk1yWVN+WPk0emNSxE8c0eWBSfVv6lFyW9O09xqfGlZxIanjFsA4nD6fP1/2P3a+lC5dmPXfpwqXad2ifb+8dlszPfOQTR7TtXdvUdmqbjk8f17v/97vVv7Vfg8sH9bPf/azsY7Dv0L7IHLfxqXF1b+5W89pmdW/uru5n0EnOuZL+Ana2duqBQw9kHYcdB3eos7UzstdP7+JeDS4fVP/Wfs0bnJe+BnoX94ZdNABAMfnSnrdJ2qmZ1rkfpL7+A0n3FNvPz42Wutp66qmnbP369XbTTTeZmdltt912UkvcW9/6VnvggQdCLml0+dVSUU/jkqpRyyUEPJUsMh0npV6jlXY9jMJx86vlqZLPM7ZnzFrWttjQzqGs907emayL6wcAUHuqsvvl6yV9T9Jjkr6smdkvX1tsPz83Ql2wpqenbd26dXbqqaeeFN6871988cV23333hVzS+PBjwoFG7gpV6zCb73y5AVe356DUa7SS6zAq165f4bLSn+dVt62ylrUt5gacLbp+EUsEAACqUnGo00wnk/MkPVfSmyS9WVJroX2C2Ah1/nriiSfsk5/8pL3tbW9LPzY7zPX29toPf/jDEEsZb35UJqPS2lFrYQSCQse6XltLy7m+4jL75ez3dAPOl9kcq/lZrNfrBwBQe9W21N1b7DlBb4S66t1www05lxY4fvy4mZl96Utfojulj6oNJt5i07MrglGdXtzPimsYYTYqLUvVKPcc1MNnzpTr87SsbbHkncms51XaDTq9PMH1i8wNuMgvTwAAqD/VhrpNkl5V7HlBboS68hw6dMiSyaRJsscff9zMTm6J6+vrsx//+Mchl7S+VbMuWqEp5qPWUud3OAhrraw4t6hUeg7i/Jlny3UzIHln0lrWtvhybXpdKb1ZLelKCQCotWpD3f2STkj6saQ9kqYk7Sm2n58boa64O+64I2dL3L/8y7+YmdnevXvtwIED4RYSJfEqp5kV9W0PbbPExkQkK5F+t6zFpdtplAJRWMcsSseg0LhIP8oYl+sSAFC/CoW6OSVMkLmihOeghn7+859r3bp12rBhg66++mp97GMf0+7du7Oes2rVKq1evVrnnXeeJKmrqyuMoqIC3nTwLc0tkqT+rf26/7H7ZTKN/eVY5KYX93v6em8Jg5GVI1q6cKl2HNyhvi19Glw+6EdxfeEtBzC7jJJCOT9hLCEQtWPQ2dqpHQd3qKejJ/3YjoM7tKhtkfZetrfq14/SMg0AAJwkX9qTNE/SByVdL+m/S5qT77lBb7TUzXSpVI6WuI6ODjObWRD8Zz/7WcilhB/i1iIQRHmj1AKUS9TOURjlidoxCHqMYNQ+LwCg8aiS7peSvirp5lSg+6ak6/I9N+it0ULdww8/bO9///tNkp155plmZnb8+PGsMPfRj37UfvGLX4RcUgQhbhNYxK28fghr3F8+YZyDqB0Ds2BvBjTidQ4AiJZCoa5Q98tFZrZYkpxzI5Lu8bmRELO0tbXp0KFDWY898cQTkqTm5mb9/ve/17x588IoGmrI67rWv7Vf+w7tU2drpwaXD0au26UnbuX1Q76ufp2tnaGUJ4xzELVjIM0ch6A+cyNe5wCA+HAzoS/HN5z7gZm9It/XtbRkyRLbtWtXGG8diAcffFCDg4O66aabJEknTpxQU1OTnHPp51x55ZW6/PLL9ZznPCesYgLII994skaq5HMMMD41rsHtg+mQm1yW5NwDQICcc/ea2ZJc3yvUUvdS59yT3mtIelbqayfJzOwMn8tZ9y6++GLdfPPNJz3+8MMPq6OjQ/kCNoBoqbbVph4qw0G1XNXDsWkEUZsoBwAaXr5+mVHa4jambmpqyv7qr/4qPf7tuuuuMzOzVatWmSR71rOeZVdffbU9+eSTIZcUaAxRmniFsVn5cWzig4ljAKD2VGBMXVO4kbI+HDt2TJL0ne98R845LV68WF/72tfS35+enpYkDQ8Py8z09NNP62Mf+5hOP/30UMoLNILxqXF1b+5W05omXfLNS3TBiy/Q4eRhDa8YVnIiqfGpcV/fp3lts7o3dxd93cHtgxpZOaKejh61NLeop6NHIytHNLg9Oks21Jp3DN/5jXfKzPTz3/2cYxNxLPEAANFCqKvQ1772NTnn5JzTKaecIkl69atfLUk6++yzdd111+npp5+WmemDH/xgmEUFYqPcgFTodZITSQ2vGFZna6euft3VGpsa0y333+JrUMh8n1IDI5XhbJnHsMk16Ya33JB1DCs5Nn5dR7V+7TjxJsrJFPZEOQDQ0PI14UVpi2L3y5e//OXp7pUXXHBB2MUBYs/PrneZXcO8qfczu4b5NfV+JV3Q6LaWLfN4eP/PPB7lHpsgu3DWS/dQP7oj5zoWbevarH19eyS6OQNAPVIl69RFaYtiqJuenrajR48WfyKAkmRW7jMrnXOvmltS5TBzHw3IRnePZr1uZpDzK0RVslZbvQQDv2QeQ+/YbHtom7kBV9GxCTI010Mg9/P6y/yZax9qt7Z1bVzXABAgQh0AM4vWhCGzeZX7zErnU0eeMjfgilYOZ1dUExsTNv+a+Ta2Zyz9vaGdQ7Zo0yJfK5uVVvKjfB7MgilfvtecfQzH9oxZYmPCNKCK3jvIRdGjuOB6uYIKpvUQeAEg6gh1ACLfQuRVCjMrh5lfl9OlcWzPmC24doElNiTs6PGjdsH4BaYBmQZkc6+aa6tuW+VLmcM6pkGGwiA+U6HX9Pv9aKkrLKhgWg+BFwCijlAHIPIVUq9y37SmyZ468lRW5b5Y5TBXhXJ096hpQOYGnLWsbbHknclAgletW92CDpJBXCfFXtPPY8iYusJoqQOA+CLUAYjFnfSxPWM296q55gZcunJfSne8QhXKeqtsBv15grhOan3tBd2SGeWus8UEFUzrIfACQNQVCnVzwpt3E0AteVOQ93T0pB+L2hTkvYt7JSk9xf0jTz6ij/2fj8nMNHrBqNrPaFfflr6s50pScllSfVv6NLJyREsXLtWOgzvUt6VPg8sH9a5b31VXSwgEvSRCENdJra+93sW96etjfGpcg9tnroPO1k4llyWzrp1qXjuOvLL3b+3XvkP71NnaqcHlg1V/pqBeFwBQonxpL0obLXVA9eJ0J91rDdGALLEhkVXGfK1SpU7EUeg14iDoz1PrMXVBitM1DwBAMaL7JQCz+HUd86PbXr1V7GvxeWo5+2WQ6i3QAwAaW6FQ52a+H21LliyxXbt2hV0MADXWvblbwyuGs7rtTR6YVP/Wfu29bG/Jr+N1wfO6hVXbBS9s9fZ5gtK8tlmHk4fV0tySfuzYiWOaNzhPJ644EWLJAAAon3PuXjNbkut7jKkDEFmFxsqVI+7joGart88TlDiMIwUAwA9NYRcAAPLpXdyrweWD6t/ar3mD89S/tT+Sky+MT42re3O3mtc2q3tzt8anxuv6fePCuykweWBSx04c0+SBSfVt6VNyWTLsogEA4Cta6gBEWr5Wqah0QRyfGldyInlSa6KkQMsT1vvGCTMyAgAaBWPqAJQkKiHKK0uuQBNGhd2vcX9xeV8AABCOQmPq6H4JoCgvRA2vGNbh5GENrxhWciJZcXe/arsNDm4f1MjKEfV09KiluUU9HT0aWTmiwe3ljbXzQ7F144LqIhn0enUAACA+CHUAivIzRPkREKMUaLzJODJ5k3H4FYZzBcNC7wsAABoLoQ5AUX6GKD8CYpQCTaHJOCr9rJkh7rz15+ny2y8/KRj2JHqYBAQAAEgi1AEogZ8hyo+AGKVZDQvN0FnJZ53dujeneY6aXbN+/rufZwXDyZ9MxmJmUAAAEDxmvwRQlF/rxUn+rB0WtVkN883QWclnzWzdk6SDTxzU7Rfdrg9990Pp9/CCIevVAQAAiVAHoAR+hqhGWlC8ks86u3XPC4CZrXuMnQMAAJkIdQBK4leIilorW5Aq+ayzW/eSy5K65JuXaOEZC3XsxLGqWkkBAEB9Yp06AHUnSmvqlSvXGnzvuOUdmjtnrn7225/F7vMAAAB/FFqnjpY6AHUl38LkkmIRhHK17l33xutiUXYAABAOWuoA1JXuzd0aXjGcNTnJ5IFJ9W/t197L9kqKd0se4oFrDADgN1rqADSMYssIxL0lD9HHNQYAqDXWqQNQV4qtqefH4udAIVxjAIBaI9QBqCvFFib3Y/FzRNv41Li6N3ereW2zujd3a3xqvKbvzzUGAKg1Qh2A0CvBhZRbtt7FvRpcPqj+rf2aNzhP/Vv7s5YRKNaSh3jzuj4OrxjW4eRhDV4dSAsAABxTSURBVK8YVnIiWdNrmmsMAFBzZhb57ZWvfKUBCMbYnjHr2NhhE/sn7Ojxozaxf8I6NnbY2J6xsIsWSNmi/HlRva5NXTaxfyLrsYn9E9a1qatmZYjLNTa2Z8y6NnVZ05om69rUFbnyAQCySdplefISE6UADS5z/I+k9Pif/q39oU/qEETZGmnx80ZUStfHoGemjMM1xmQuAFBfCHVAg4vy+J+gyta7uJeKa53yuj5mLmmR2fWxVmEm6tdYlG/mAADKx5g6oMFFefxPlMuGaCo2UQ4zU86I8s0cAED5Qgl1zrmznHO3OOcecM7tc879cRjlAFC8EkzZECfFJsqJY5gJYiIjbpgAQH0Jq/vldZJuN7MLnXOnSDo1pHIADS/X+J83/eGbNLh9UO+69V2BjDmqpmxRG5uE6CnU9bFY98yoCaq7qHfDZPbrDi5vrBZLAKgXbmYilRq+oXNnSLpP0gusxDdfsmSJ7dq1K9iCAZCUvxJJmCou6Ak4UL24Xd/dm7s1vGI4K4ROHphU/9Z+7b1sb1WvzfUKAPHinLvXzJbk/F4Ioe5lkm6QdL+kl0q6V9LlZvbUrOe9T9L7JGnhwoWvfPjhh2taTqBRBVmJrGdxCwuNLEphplhZmtc263DysFqaW9KPHTtxTPMG5+nEFSfCKDIAICSFQl0Y3S/nSHqFpH4zu9s5d52kj0v658wnmdkNmgl/WrJkSW2TJ9DA4jjmKAqYTTA+ojIzZSldK+PWXRQAEI4wJkp5RNIjZnZ36utbNBPyAEQAEyhUhjCMcpUyEyeTBQEASlHzljoz+7lz7j+dcy8ysx9J+nPNdMUEEAFMoFAZWlRQrlJuBDBZEACgFGHNftkv6cupmS/3S/q7kMoBYBYqkZUhDKNcpd4IiEp3UQBAdIUS6sxst6Scg/wAhI9KZPkIwygXNwIAAH6p+eyXlWBJA6CxVTpbYZRmOQRy4RoFAJQqarNfApiFil1+lS6+HNSizYCfaBUHAPghjNkvgdCMT42re3O3mtc2q3tzt8anxsMuUjp8DK8Y1uHkYQ2vGFZyIhmJskVBKTME+rkfAABA3BDq0DCiGp4IH4VVulRAVJcYiOKNhajjmAEAUBihDg0jquEpquEjKipdNy+K6+1F9cZClHHMgOpwUwRoDIQ6NIyohqcoho8oqXTx5dn7fWLiE3rDzW/Q/Y/dH1rFJqo3FqKMYwZUjpsiQONgohQ0jKguDs205oVVulRA5n73P3a/5jTN0eo/Xa0r/+zK0CZNieqNhSjjmAGVy7wpIil9U6R/az8T9AB1hpY6NIxKW3yC1ru4V4PLB9W/tV/zBuepf2t/Xa9vVsuuQL2Le7X3sr1a1LZI333Xd/XJ5Z8MtbWHVtnyccyAynFTBGgctNShYUR5cehGmda8kmUG/FiaICoVG1ply8cxA8qTuUROS1OL1nxvjT65/JPp73NTBKhTZhb57ZWvfKUBiL+uTV02sX8i67GJ/RPWtanL132CeA2/jO0Zs65NXda0psm6NnXZ2J6xmpchbjhmQGnG9oxZx8YOm9g/YUePH7XknUmbs3aOJe9M2tHjR21i/4R1bOzgZwiIKUm7LE9ecjPfj7YlS5bYrl27wi4GgCo1r23W4eRhtTS3pB87duKY5g3O04krTvi2z2z5Wvui0lILAH7o3tyt4RXDWWPHPzHxCV2781odmz6mztZOJZcl+b0HxJRz7l4zW5Lre3S/BFAzlUxW48cEN7XsepvZ9YkKFIBaytXV/Mo/u1Kf3vHpkm+CAYgnQh2AmqlkfJRfY6pqMW7Rj/F/AFCpqM7yDCB4hDoANVNJi1mUJ7iZjenDAYSJiYWAxkWoAxCoXN0R9162t6zXiMvsoFGZZRNAY4rTTTAA/iLUAQhMo3VHpOsTgLDF5SYYAH+x+DiAwGR2Rwxz0e9aieoC9wAAoL7RUgcgMI3WHZGuTwAAIAyEOgCBCas7YpjLCtD1CQAA1BrdLwEEJozuiN44vuEVwzqcPKzhFcNKTiQ1PjUe2HsCAACEiZY6AIEJozsiywoAAIBGQ6gDEKhad0dstHF8AAAAdL8EUFe8cXyZWFYAAADUM0IdgLrCsgIAAKDR0P0SQF1hWQEAANBonJmFXYailixZYrt27Qq7GAAAAAAQCufcvWa2JNf36H4JAAAAADFGqAPQUManxtW9uVvNa5vVvbmb9esAAEDsMaYOQMPwFiYfWTmipQuXasfBHerb0idJjLkDAACxRUsdEBBahKInc2HyluaW9MLkg9sHwy4aAABAxQh1QAC8FqHhFcM6nDys4RXDSk4kCXYhK7YwOUEcAADEEaEOCAAtQtFUaGFygjgAAIgrQh0QgGItQigsqBazQguTE8QBAEBcMVEKEACvRainoyf9mNcihMKCnMyk0MLk77r1XXUVxMenxjW4fTD9OZPLkkwGAwBAnaKlDghAoRahWovbOLGgW8x6F/dq72V7deKKE9p72d500CnUNTNu6EoKAEBjoaUOCEChFqFaiuMU/mF1XfWC+OxjNbg8ft0vM4OxpHQw7t/aH9nzDgAAKkeoAwLSu7g39Ap0HCv3YXVdjUoQ9wNjOgEAaCyEOqCOxbFyH2aLWRSCuB8Y0wkAQGMh1AF1LI6V+3pqMQtLPXUlBQAAxRHqgDqQb6bDWlfu/ZpxsV5azMJCMAYAoLEQ6oCYK2UylFpU7uM4KUs9IxgDANA4nJmFXYailixZYrt27Qq7GEAkdW/u1vCK4awulpMHJtW/tV97L9vbcOUIGuu/AQCAMDjn7jWzJbm+R0sdEHNRmQwlKuUIEq2RAAAgilh8HIi5qCyaHZVyBCnohdEBAAAqQagDYs6bDGXywKSOnTimyQOT6tvSp+SyZEOWI0iN0BoJAADih+6XQMxFZabDqJQjSHFcIgIAANQ/JkoB6giTeAQr35i6eguvAAAgepgoBWgATOIRvEZojQQAAPFDSx1QJxplSQEAAIBGVKiljolSgDrBJB4AAACNiVAH1IlGWFKgVONT4+re3K3mtc3q3tyt8anxsIsEAAAQGMbUAXVgfGpcTxx+QstHlytxZkJre9aq/Yz29CQejYSxhQAAoNEwpg6IucwQ88iTj+iKu67QT37zE7Wf0a51r1vXcEGGsYUAAKAeMaYOqGOD2wc1snJEPR09uvilF+vA5Qc08bcTOnPumYEHuih2c2RsIQAAaDSEOiAiKg1IYYUYr4VweMWwDicPa3jFsJITydCDHWMLAQBAoyHUARFQTUAKK8RkthC2NLeop6NHIytHNLg93DF8yWVJ9W3p0+SBSR07cUyTBybVt6VPyWXJUMsFAAAQFCZKASIgMyBJSgek/q39RbtQeiFm9sQgQU+QEtVujiwQDgAAGg2hDoiAagJSWCHGayHMnJAkKt0cexf3EuIAAEDDINQBEVBtQAojxITVQggAAIBsjKkDIiCO48B6F/dqcPmg+rf2a97gPPVv7a+4hTCKs2gCAADEBS11QATEdRyYHy2ELBYOAABQHRYfBxAqFgsHAAAojsXHAURWVGfRBAAAiAtCHYBQsVg4AABAdQh1QECY/KM0cZwkBgAAIEqYKAUIwPjUuC6//XKd1nKazExPHX1Kl99+uSQm/5gtrpPEAAAARAUTpQABOG/DeTp+4rjG3j6WntHxnV9/p+Y0z9F/fug/wy4eAAAAYoaJUoAae+TJRzT6tlH1dPSopblFPR09Gn3bqB558pGwiwYAAIA6Q6gDAAAAgBgj1AEBaD+9XZd885KsyT8u+eYlaj+9PeyiAQAAoM4Q6oAArHv9Oh2fPq73bHmP5g3O03u2vEfHp49r3evXhV00AAAA1BlCHRCA3sW9uu6N1+m0ltMkSae1nKbr3ngdMzoCAADAd4Q6ACdhjT0AAID4YJ06IADjU+NKTiQ1snIkvaRB35Y+SdFfpy7OZQcAAGhEtNQBARjcPqiRlSNZSxqMrBzR4PbBsItWVJzLDgAA0IgIdUAA9h3ap6ULl2Y9tnThUu07tC+kEpUuzmUHAABoRIQ6IACdrZ3acXBH1mM7Du5QZ2tnSCUqXZzLDgAA0IgIdUAAksuS6tvSl7VOXd+WPiWXJcMuWlFxLjsAAEAjYqIUIADehCL9W/u179A+dbZ2anD5YCwmGolz2QEAABqRM7Owy1DUkiVLbNeuXWEXAwAAAABC4Zy718yW5Poe3S8BAAAAIMYIdQAAAAAQY4Q6AAAAAIixUCZKcc79RNJvJZ2QdDxf31AAAAAAQGFhttT1mNnLCHSoV+NT4+re3K3mtc3q3tyt8anxsIsEAACAOkT3SyAA41PjSk4kNbxiWIeThzW8YljJiSTBDrHFTQoAAKIrrFBnkrY55+51zr0v1xOcc+9zzu1yzu167LHHalw8oDqD2wc1snJEPR09amluUU9Hj0ZWjmhw+2DYRQPKxk0KAACiLZR16pxz55jZz5xzz5N0h6R+M/vXfM9nnTrETfPaZh1OHlZLc0v6sWMnjmne4DyduOJEiCUDyte9uVvDK4bV09GTfmzywKT6t/Zr72V7QywZAACNI3Lr1JnZz1L//lLSrZL+KIxyAEHpbO3UjoM7sh7bcXCHOls7QyoRULl9h/Zp6cKlWY8tXbhU+w7tC6lEAAAgU81DnXPuNOfc6d7/Jf2FJG71oq4klyXVt6VPkwcmdezEMU0emFTflj4llyXDLhpQNm5SAAAQbWEsafB8Sbc657z3HzOz20MoBxCY3sW9kqT+rf3ad2ifOls7Nbh8MP04ECfeTYqRlSNaunCpdhzcob4tfRpczhhRAACiIJQxdeViTB0AhGt8alyD2wfTNymSy5LcpAAAoIYKjakj1AEAAABAxEVuohQAAAAAgD8IdQAAAAAQY4Q6AAAAAIgxQh0AAAAAxBihDgAAAABijFAHAAAAADFGqAMAAACAGCPUAQAAAECMEeoAAAAAIMYIdQAAAAAQY4Q6AAAAAIgxQh0AAAAAxBihDgAAAABijFAHoCTjU+Pq3tyt5rXN6t7crfGp8bCLBAAAAElzwi4AgOgbnxpXciKpkZUjWrpwqXYc3KG+LX2SpN7FvSGXDgAAoLHRUgc0uFJa4Aa3D2pk5Yh6OnrU0tyino4ejawc0eD2wRBKDAAAgEy01AENrNQWuH2H9mnpwqVZ+y5duFT7Du2raXkBAABwMlrqgAZWagtcZ2undhzckfXYjoM71NnaWcviAgAAIAdCHdDASm2BSy5Lqm9LnyYPTOrYiWOaPDCpvi19Si5L1rK4AAAAyIHul0AD81rgejp60o/laoHzumL2b+3XvkP71NnaqcHlg0ySAgAAEAGEOqCBeS1ws8fUDS4/eQKU3sW9hDgAAIAIItQBDYwWOAAAgPhzZhZ2GYpasmSJ7dq1K+xiAAAAAEAonHP3mtmSXN9johQAAAAAiDFCHQAAAADEGKEOAAAAAGKMUAcAAAAAMUaoAwAAAIAYI9QBAAAAQIwR6iowPjWu7s3dal7brO7N3RqfGg+7SAAAAAAaFKGuTONT40pOJDW8YliHk4c1vGJYyYkkwQ4AUDFuFgIAqkGoK9Pg9kGNrBxRT0ePWppb1NPRo5GVIxrcPhh20QAAMcTNQgBAtQh1Zdp3aJ+WLlya9djShUu179C+kEoEAIgzbhYCAKpFqCtTZ2undhzckfXYjoM71NnaGVKJAABxxs1CAEC1CHVlSi5Lqm9LnyYPTOrYiWOaPDCpvi19Si5Lhl00AEAMcbMQAFCtOWEXIG56F/dKkvq39mvfoX3qbO3U4PLB9OMAAJTDu1k4snJESxcu1Y6DO9S3pU+Dy+l+CQAojTOzsMtQ1JIlS2zXrl1hFwMAgECMT41rcPtg+mZhclmSm4UAgCzOuXvNbEnO7xHqAAAAACDaCoU6xtQBAAAAQIwR6gAAAAAgxgh1AAAAABBjhDoAAAAAiDFCHQAAAADEGKEOAAAAAGKMUAcAAAAAMUaoAwAAAIAYI9QBAAAAQIwR6gAAAAAgxgh1AAAAABBjhDoAAAAAiDFCHQAAAADEGKEOAAAAAGKMUAcAAAAAMUaoAwAAAIAYI9QBAAAAQIwR6gAAAAAgxgh1AAAAABBjhDoAAAAAiDFnZmGXoSjn3GOSHg67HHWsVdKhsAuBvDg/0cb5iS7OTbRxfqKN8xNtnJ/oCvLcnG9mbbm+EYtQh2A553aZ2ZKwy4HcOD/RxvmJLs5NtHF+oo3zE22cn+gK69zQ/RIAAAAAYoxQBwAAAAAxRqiDJN0QdgFQEOcn2jg/0cW5iTbOT7RxfqKN8xNdoZwbxtQBAAAAQIzRUgcAAAAAMUaoAwAAAIAYI9Q1MOfcec65SefcPufcD51zl4ddJmRzzjU75/7dOfftsMuCbM65s5xztzjnHkj9DP1x2GXCM5xzH0r9XtvrnBt3zs0Lu0yNzDl3o3Pul865vRmPne2cu8M592Dq3+eEWcZGluf8XJP6/bbHOXerc+6sMMvYqHKdm4zvfcQ5Z8651jDKhvznxznX75z7Uerv0LpalIVQ19iOS/ofZtYp6TWS/sE5tyjkMiHb5ZL2hV0I5HSdpNvN7MWSXirOU2Q4586V9AFJS8ysW1KzpL8Jt1QN70uS3jjrsY9LutPM/lDSnamvEY4v6eTzc4ekbjN7iaT/kPSPtS4UJOU+N3LOnSfp9ZIO1rpAyPIlzTo/zrkeSW+V9BIz65J0bS0KQqhrYGb2qJn9IPX/32qmUnpuuKWCxznXLulNkr4QdlmQzTl3hqT/KmlEkszsqJn9JtxSYZY5kp7lnJsj6VRJPwu5PA3NzP5V0q9nPfxWSTel/n+TpAtqWiik5To/ZrbNzI6nvvy/ktprXjDk+9mRpA2SVktixsMQ5Tk/l0q62syOpJ7zy1qUhVAHSZJzLiHp5ZLuDrckyLBRM7+wp8MuCE7yAkmPSfpiqnvsF5xzp4VdKMwws59q5s7oQUmPSnrCzLaFWyrk8Hwze1Saucko6Xkhlwf5vUfS1rALgRnOuZWSfmpm94VdFuT0QknLnHN3O+e+55x7VS3elFAHOeeeLenrkj5oZk+GXR5Izrk3S/qlmd0bdlmQ0xxJr5D0OTN7uaSnRNexyEiNzXqrpA5J50g6zTn3rnBLBcSTcy6pmeEaXw67LJCcc6dKSkq6IuyyIK85kp6jmaFNH5X0/znnXNBvSqhrcM65Fs0Eui+b2TfCLg/S/lTSSufcTyR9RdJy59zN4RYJGR6R9IiZeS3bt2gm5CEaXifpgJk9ZmbHJH1D0p+EXCac7BfOuQWSlPq3Jl2UUDrn3CWS3izpImNh46j4A83csLovVUdol/QD59z8UEuFTI9I+obNuEczPa4Cn8yGUNfAUncNRiTtM7P1YZcHzzCzfzSzdjNLaGaChwkzo6UhIszs55L+0zn3otRDfy7p/hCLhGwHJb3GOXdq6vfcn4uJbKJoi6RLUv+/RNL/DrEsmMU590ZJH5O00syeDrs8mGFmU2b2PDNLpOoIj0h6RervEqLhm5KWS5Jz7oWSTpF0KOg3JdQ1tj+VdLFmWoF2p7b/FnahgJjol/Rl59weSS+T9KmQy4OUVAvqLZJ+IGlKM3/rbgi1UA3OOTcu6d8kvcg594hzrk/S1ZJe75x7UDOz+F0dZhkbWZ7zc72k0yXdkaof/EuohWxQec4NIiLP+blR0gtSyxx8RdIltWjpdrSmAwAAAEB80VIHAAAAADFGqAMAAACAGCPUAQAAAECMEeoAAAAAIMYIdQAAAAAQY4Q6AEBezrm3OefMOffiEp77bufcOVW812udc9/O87rX59lnZ6Xv4Zxb6Zz7eGWlLfn92pxzdzvn/t05t8w591fOuX3OuUnn3BLn3GdLeI2dqX8Tzrl3Vlmeu5xzS4o854POuVOreR8AQG0R6gAAhfRK2iHpb0p47rslVRzqKmFmf1LFvlvMLOi10f5c0gNm9nIz2y6pT9JlZtZjZrvM7AMllNP7jAlJVYW6En1QEqEOAGKEUAcAyMk592xJf6qZIPI3s7632jk35Zy7zzl3tXPuQklLNLMg+27n3LOccz9xzrWmnr/EOXdX6v9/5JzbmWq92umce1EJxTnPOXe7c+5HzrkrM8rxu9S/r021Qt3inHvAOfdl55xLfe+Nqcd2SPrLjH3TLYDOuS855z6bKs/+1OeRc67JObfZOfdD59y3nXPf8b4363j8vXPu/6WOx9edc6c6514maZ2k/5Y6JldKWirpX5xz18xqNRxwzt2Y+gz7nXMfyHjt36X+e7WkZanX+pBzbnvqPbznfd8595JZ5XqWc+4rzrk9zrmvSnpWxvc+55zblfpsa1KPfUAzwXzSOTeZ73kAgGiZE3YBAACRdYGk283sP5xzv3bOvcLMfuCcW5H63qvN7Gnn3Nlm9mvn3CpJHzGzXZKUylS5PCDpv5rZcefc6yR9StLbi5TljyR1S3pa0v9zzt3mvU+Gl0vqkvQzSd+X9KfOuV2SPi9puaSHJH21wHss0EzoerGkLZJu0UwITEhaLOl5kvZJujHHvt8ws89LknPuk5L6zGzYOXeFpCVmtir1vR6ljpFz7rWzXuPFknoknS7pR865z5nZsYzvfzy175tTr/VrzbSOftA590JJc81sz6zXvFTS02b2klTg+0HG95Kp89Ys6U7n3EvM7LPOuQ9L6jGzQwWeN/t9AAAhoqUOAJBPr6SvpP7/ldTXkvQ6SV80s6clycx+Xebrninpa865vZI2aCaIFXOHmf3KzH4v6RuaCV+z3WNmj5jZtKTdmgljL5Z0wMweNDOTdHOB9/immU2b2f2Snp96bKmkr6Ue/7mkyTz7dqdazqYkXVTiZ5rtNjM7kgpTv8woQz5fk/Rm51yLpPdI+lKO5/xXpT5zKohlhrG/ds79QNK/p8q7KM/7lPo8AEBIaKkDAJzEOfdczbRudTvnTFKzJHPOrZbkJFkJL3Ncz9w8nJfx+FWSJs3sbc65hKS7Snit2e+X6/2PZPz/hJ75G1dKWWfv72b9W8yXJF1gZvc5594t6bUl7pfv/TPLn1OqlfQOSW+V9Nea6f6a86mzH3DOdUj6iKRXmdnjzrkvKfsclfU8AEC4aKkDAORyoaRRMzvfzBJmdp6kA5ppudom6T3eDInOubNT+/xWM10HPT+R9MrU/zO7V54p6aep/7+7xPK83jl3tnPuWZrp+vn9Evd7QFKHc+4PUl/3FnpyDjskvT01tu75yh/WTpf0aKrV7KIy36NUs4+vJH1B0mcl/b88Lab/6pXHOdctyRtzd4akpyQ9kfpcK/K8T6HnAQAiglAHAMilV9Ktsx77uqR3mtntmhlztss5t1szLTnSTGvVv3gTpUhaI+k659x2zbQ8edZJ+rRz7vuaaQEsxQ5J/0sz3Sq/nmM8XU5mdljS+yTdlpoo5eES38/zdUmPSNor6X9KulvSEzme98+p792hmSAZhD2SjqcmY/mQJJnZvZKelPTFPPt8TtKznXN7JK2WdE9qv/s0053yh5oZI5gZkm+QtNU5N1nkeQCAiHAzQwwAAEAuzrlnm9nvUl1S75H0p6nxdaFzM+sC3iXpxamxhACABsSYOgAACvu2c+4sSadIuipCge5vJQ1K+jCBDgAaGy11AAAAABBjjKkDAAAAgBgj1AEAAABAjBHqAAAAACDGCHUAAAAAEGOEOgAAAACIsf8f+sdy01p+HHsAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 1080x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "\n",
    "fig_size = np.zeros(2)\n",
    "fig_size[0] = 15\n",
    "fig_size[1] = 10\n",
    "plt.rcParams[\"figure.figsize\"] = fig_size\n",
    "\n",
    "f, ax = plt.subplots()\n",
    "\n",
    "plt.plot(y_test, preds, 'o', markerfacecolor=\"None\", markeredgecolor='green', markeredgewidth=1)\n",
    "#ax.plot(ax.get_xlim(), ax.get_ylim(), ls=\"--\", c=\".3\")\n",
    "coef = np.polyfit(y_test,preds,1)\n",
    "poly1d_fn = np.poly1d(coef)\n",
    "plt.plot(y_test, poly1d_fn(y_test), '--k')\n",
    "\n",
    "plt.title(\"Test data binding affinity\")\n",
    "plt.xlabel('Actual binding affinity data')\n",
    "plt.ylabel('Predicted binding affinity data')\n",
    "#plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Top Feature ranking:\n",
      "1. feature 0 (0.111758)\n",
      "2. feature 2 (0.103460)\n",
      "3. feature 8 (0.102103)\n",
      "4. feature 4 (0.096238)\n",
      "5. feature 9 (0.088906)\n",
      "6. feature 6 (0.083047)\n",
      "7. feature 10 (0.077822)\n",
      "8. feature 1 (0.077635)\n",
      "9. feature 5 (0.060353)\n",
      "10. feature 12 (0.042364)\n",
      "11. feature 3 (0.041709)\n",
      "12. feature 13 (0.026116)\n",
      "13. feature 14 (0.023586)\n",
      "14. feature 7 (0.021648)\n",
      "15. feature 11 (0.019878)\n",
      "16. feature 15 (0.017996)\n",
      "17. feature 16 (0.002876)\n",
      "18. feature 17 (0.002505)\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA34AAAJdCAYAAABtZqU4AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjAsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+17YcXAAAgAElEQVR4nOzdebhkd10n/veHBEIkECQJazKEJaBhEaFZfjOoSBRBgYATfizKoiioLIOIiowLojigMIwOoMIECbiEgIphDQqCCgrpMGEJEOiEYGJQAoQlbCHhM3/UaXJzuZ2uTu651f3N6/U89+k6p75V9a7T59a973u26u4AAAAwrmusOgAAAADzUvwAAAAGp/gBAAAMTvEDAAAYnOIHAAAwOMUPAABgcIofAOxCVf1RVf3aqnMAwFVVruMHwGarqnOS3CjJpWtm36a7z78Kz3mvJH/a3YdftXT7pqp6eZLzuvtXV50FgH2PLX4AzOUB3X3Qmq8rXfo2Q1Xtv8rXvyqqar9VZwBg36b4AbClquoeVfWuqvpcVb1v2pK3876fqKoPV9UXq+rsqnr8NP86Sd6U5KZVddH0ddOqenlV/faax9+rqs5bM31OVf1yVb0/yZeqav/pcX9ZVRdU1cer6slXkPWbz7/zuavql6rqU1X1yap6UFX9cFV9tKo+W1XPWPPYZ1bVa6rqVdP7eW9Vfdea+7+zqt4+LYczquqB6173D6vqjVX1pSSPTfJjSX5peu+vm8Y9varOmp7/Q1X14DXP8Ziq+qeqel5VXTi91/utuf8GVfUnVXX+dP9r19x3/6o6fcr2rqq645r7frmq/m16zTOr6pgl/tsBWDHFD4AtU1U3S/KGJL+d5AZJnpbkL6vqsGnIp5LcP8n1kvxEkhdU1Z27+0tJ7pfk/CuxBfHhSX4kyfWTfCPJ65K8L8nNkhyT5ClV9UNLPteNk1x7euyvJ3lpkh9Pcpck35Pk16vqlmvGH5vk1dN7/fMkr62qa1bVNaccb0lywyRPSvJnVXXbNY99RJJnJ7luklck+bMkvzu99wdMY86aXvfgJL+Z5E+r6iZrnuPuSc5McmiS301yfFXVdN8rk3xbkttNGV6QJFV15yQvS/L4JIck+eMkJ1fVAVO+Jya5a3dfN8kPJTlnyWUHwAopfgDM5bXTFqPPrdma9ONJ3tjdb+zub3T33ybZnuSHk6S739DdZ/XCO7IoRt9zFXP8QXef291fSXLXJId197O6++LuPjuL8vawJZ/r60me3d1fT3JiFoXq97v7i919RpIzktxxzfjTuvs10/j/mUVpvMf0dVCS50w53pbk9VmU1J3+prvfOS2nr24Uprtf3d3nT2NeleRjSe62Zsgnuvul3X1pkhOS3CTJjaZyeL8kP9PdF3b316flnSQ/neSPu/vd3X1pd5+Q5GtT5kuTHJDk6Kq6Znef091nLbnsAFghxQ+AuTyou68/fT1omnfzJA9ZUwg/l+SeWRSSVNX9qupfpt0mP5dFITz0KuY4d83tm2exu+ja139GFieiWcZnphKVJF+Z/v2PNfd/JYtC9y2v3d3fSHJekptOX+dO83b6RBZbEjfKvaGqetSaXTI/l+T2ufzy+vc1r//l6eZBSY5I8tnuvnCDp715kl9Yt4yOSHLT7t6R5ClJnpnkU1V1YlXddHc5AVg9xQ+ArXRukleuKYTX7+7rdPdzquqAJH+Z5HlJbtTd10/yxiQ7d03c6DTUX8pid8WdbrzBmLWPOzfJx9e9/nW7+4ev8jvb2BE7b1TVNZIcnuT86euIad5O/ynJv+0i97dMV9XNs9ha+cQkh0zL64O5bHldkXOT3KCqrr+L+569bhl9W3f/RZJ095939z2zKIid5LlLvB4AK6b4AbCV/jTJA6rqh6pqv6q69nTSlMOTXCuL3QgvSHLJdCKS+6x57H8kOaSqDl4z7/QkPzydqOTGWWyNuiLvSfKF6QQlB04Zbl9Vd920d3h5d6mqH63FGUWfksUuk/+S5N1ZlNZfmo75u1eSB2Sx++iu/EeStccPXieL4nVBsjgxThZb/Haruz+ZxclyXlxV3z5l+N7p7pcm+ZmqunstXKeqfqSqrltVt62qe08l/atZbOG8dBcvA8BeRPEDYMt097lZnPDkGVkUlnOT/GKSa3T3F5M8OclJSS7M4uQmJ6957EeS/EWSs6ddEG+axQlK3pfFCUbekuRVu3n9S7MoWHdK8vEkn07yf7I4Ococ/ibJQ7N4P49M8qPT8XQXJ3lgFsfZfTrJi5M8anqPu3J8FsfWfa6qXtvdH0ry/CT/nEUpvEOSd+5BtkdmccziR7I4qc5TkqS7t2dxnN8Lp9w7kjxmeswBSZ4zZf73LE4K84wAsNdzAXcAmEFVPTPJrbv7x1edBQBs8QMAABic4gcAADA4u3oCAAAMzhY/AACAwe2/6gCb5dBDD+0jjzxy1TEAAABW4rTTTvt0dx+20X3DFL8jjzwy27dvX3UMAACAlaiqT+zqPrt6AgAADE7xAwAAGJziBwAAMDjFDwAAYHCKHwAAwOAUPwAAgMEpfgAAAINT/AAAAAan+AEAAAxO8QMAABic4gcAADA4xQ8AAGBwih8AAMDgFD8AAIDBKX4AAACDU/wAAAAGp/gBAAAMTvEDAAAYnOIHAAAwOMUPAABgcIofAADA4BQ/AACAwSl+AAAAg1P8AAAABrf/qgPsk6pWneAy3atOAAAA7OVs8QMAABic4gcAADA4xQ8AAGBwih8AAMDgFD8AAIDBKX4AAACDU/wAAAAGp/gBAAAMTvEDAAAYnOIHAAAwOMUPAABgcIofAADA4BQ/AACAwSl+AAAAg1P8AAAABqf4AQAADE7xAwAAGJziBwAAMDjFDwAAYHCKHwAAwOAUPwAAgMEpfgAAAINT/AAAAAan+AEAAAxO8QMAABic4gcAADA4xQ8AAGBwih8AAMDgFD8AAIDBKX4AAACDU/wAAAAGp/gBAAAMTvEDAAAYnOIHAAAwOMUPAABgcIofAADA4BQ/AACAwSl+AAAAg9t/1QHYAlWrTnCZ7lUnAACAqx1b/AAAAAan+AEAAAxO8QMAABic4gcAADA4xQ8AAGBwzurJ3sdZSAEAYFPZ4gcAADA4xQ8AAGBwih8AAMDgFD8AAIDBKX4AAACDU/wAAAAGN2vxq6r7VtWZVbWjqp6+wf3fW1XvrapLquq4dfc9uqo+Nn09es6cAAAAI5ut+FXVfklelOR+SY5O8vCqOnrdsH9N8pgkf77usTdI8htJ7p7kbkl+o6q+fa6sAAAAI5tzi9/dkuzo7rO7++IkJyY5du2A7j6nu9+f5BvrHvtDSf62uz/b3Rcm+dsk950xKwAAwLDmLH43S3Lumunzpnmb9tiqelxVba+q7RdccMGVDgoAADCyOYtfbTCvN/Ox3f2S7t7W3dsOO+ywPQoHAABwdTFn8TsvyRFrpg9Pcv4WPBYAAIA15ix+pyY5qqpuUVXXSvKwJCcv+dhTktynqr59OqnLfaZ5AAAA7KHZil93X5LkiVkUtg8nOam7z6iqZ1XVA5Okqu5aVecleUiSP66qM6bHfjbJb2VRHk9N8qxpHgAAAHuoupc97G7vtm3btt6+ffvWvFhtdAjiiizz/yfvlTfI9wcAAOOrqtO6e9tG9816AXcAAABWT/EDAAAYnOIHAAAwOMUPAABgcIofAADA4BQ/AACAwSl+AAAAg1P8AAAABqf4AQAADE7xAwAAGJziBwAAMLj9Vx0A9nlVq05wme5VJwAAYC9kix8AAMDgFD8AAIDBKX4AAACDU/wAAAAGp/gBAAAMTvEDAAAYnOIHAAAwOMUPAABgcIofAADA4BQ/AACAwSl+AAAAg1P8AAAABqf4AQAADE7xAwAAGJziBwAAMDjFDwAAYHCKHwAAwOAUPwAAgMEpfgAAAINT/AAAAAan+AEAAAxO8QMAABic4gcAADA4xQ8AAGBwih8AAMDgFD8AAIDBKX4AAACDU/wAAAAGp/gBAAAMTvEDAAAYnOIHAAAwOMUPAABgcIofAADA4PZfdQBgi1WtOsFluledAADgasEWPwAAgMEpfgAAAINT/AAAAAan+AEAAAxO8QMAABic4gcAADA4xQ8AAGBwih8AAMDgFD8AAIDBKX4AAACDU/wAAAAGp/gBAAAMTvEDAAAYnOIHAAAwOMUPAABgcIofAADA4BQ/AACAwSl+AAAAg1P8AAAABqf4AQAADE7xAwAAGJziBwAAMDjFDwAAYHCKHwAAwOAUPwAAgMEpfgAAAIPbf9UBAK5Q1aoTXKZ71QkAAK4UW/wAAAAGp/gBAAAMTvEDAAAYnOIHAAAwOMUPAABgcIofAADA4BQ/AACAwSl+AAAAg1P8AAAABqf4AQAADG7W4ldV962qM6tqR1U9fYP7D6iqV033v7uqjpzmX7OqTqiqD1TVh6vqV+bMCQAAMLLZil9V7ZfkRUnul+ToJA+vqqPXDXtskgu7+9ZJXpDkudP8hyQ5oLvvkOQuSR6/sxQCAACwZ+bc4ne3JDu6++zuvjjJiUmOXTfm2CQnTLdfk+SYqqokneQ6VbV/kgOTXJzkCzNmBQAAGNacxe9mSc5dM33eNG/DMd19SZLPJzkkixL4pSSfTPKvSZ7X3Z9d/wJV9biq2l5V2y+44ILNfwcAAAADmLP41Qbzeskxd0tyaZKbJrlFkl+oqlt+y8Dul3T3tu7edthhh13VvAAAAEOas/idl+SINdOHJzl/V2Om3ToPTvLZJI9I8ubu/np3fyrJO5NsmzErAADAsOYsfqcmOaqqblFV10rysCQnrxtzcpJHT7ePS/K27u4sdu+8dy1cJ8k9knxkxqwAAADDmq34TcfsPTHJKUk+nOSk7j6jqp5VVQ+chh2f5JCq2pHkqUl2XvLhRUkOSvLBLArkn3T3++fKCgAAMLJabGDb923btq23b9++NS9WGx2auCLL/P/Je+XJO68R8wIArEhVndbdGx4iN+sF3AEAAFg9xQ8AAGBwih8AAMDgFD8AAIDB7b/qAABDcTIaAGAvpPgBXJ0pqgBwtWBXTwAAgMEpfgAAAINT/AAAAAan+AEAAAxO8QMAABic4gcAADA4xQ8AAGBwih8AAMDgFD8AAIDBKX4AAACDU/wAAAAGp/gBAAAMTvEDAAAYnOIHAAAwOMUPAABgcIofAADA4BQ/AACAwSl+AAAAg1P8AAAABqf4AQAADE7xAwAAGJziBwAAMDjFDwAAYHCKHwAAwOAUPwAAgMEpfgAAAINT/AAAAAan+AEAAAxO8QMAABic4gcAADA4xQ8AAGBwih8AAMDgFD8AAIDBKX4AAACDU/wAAAAGp/gBAAAMTvEDAAAYnOIHAAAwOMUPAABgcIofAADA4BQ/AACAwSl+AAAAg1P8AAAABqf4AQAADE7xAwAAGJziBwAAMLjdFr+qulFVHV9Vb5qmj66qx84fDQAAgM2wzBa/lyc5JclNp+mPJnnKXIEAAADYXMsUv0O7+6Qk30iS7r4kyaWzpgIAAGDTLFP8vlRVhyTpJKmqeyT5/KypAAAA2DT7LzHmqUlOTnKrqnpnksOSHDdrKgAAADbNbotfd7+3qr4vyW2TVJIzu/vrsycDAABgU+y2+FXVo9bNunNVpbtfMVMmAAAANtEyu3redc3tayc5Jsl7kyh+AAAA+4BldvV80trpqjo4yStnSwQAAMCmWuasnut9OclRmx0EAACAeSxzjN/rMl3KIYuieHSSk+YMBQAAwOZZ5hi/5625fUmST3T3eTPlAQAAYJMtc4zfO7YiCAAAAPPYZfGrqi/msl08L3dXku7u682WCgAAgE2zy+LX3dfdyiAAAADMY5lj/JIkVXXDLK7jlyTp7n+dJREAAACbareXc6iqB1bVx5J8PMk7kpyT5E0z5wIAAGCTLHMdv99Kco8kH+3uWyQ5Jsk7Z00FAADAplmm+H29uz+T5BpVdY3u/vskd5o5FwAAAJtkmWP8PldVByX5hyR/VlWfyuJ6fgAAAOwDltnid2ySLyf5+SRvTnJWkgfMGQoAAIDNs8wWv8cleXV3n5fkhJnzAAAAsMmW2eJ3vSSnVNU/VtUTqupGc4cCAABg8+y2+HX3b3b37ZI8IclNk7yjqv5u9mQAAABsimW2+O30qST/nuQzSW44TxwAAAA22zIXcP/Zqnp7krcmOTTJT3f3HecOBgAAwOZY5uQuN0/ylO4+fe4wAAAAbL7dFr/ufvpWBAEAAGAee3KMHwAAAPsgxQ8AAGBwSxW/qrp5Vf3AdPvAqrruvLEAAADYLMuc1fOnk7wmyR9Psw5P8to5QwEAALB5ltni94Qk/yXJF5Kkuz+WJa/jV1X3raozq2pHVX3LSWKq6oCqetV0/7ur6sg1992xqv65qs6oqg9U1bWXeU0AAAAub5nLOXytuy+uqiRJVe2fpHf3oKraL8mLkvxgkvOSnFpVJ3f3h9YMe2ySC7v71lX1sCTPTfLQ6TX+NMkju/t9VXVIkq/vyRsDYEDTz6K9Qu/2RyEA7DWW2eL3jqp6RpIDq+oHk7w6yeuWeNzdkuzo7rO7++IkJyY5dt2YY5OcMN1+TZJjatEw75Pk/d39viTp7s9096VLvCYAAADrLFP8np7kgiQfSPL4JG9M8qtLPO5mSc5dM33eNG/DMd19SZLPJzkkyW2SdFWdUlXvrapf2ugFqupxVbW9qrZfcMEFS0QCAAC4+llmV88Dk7ysu1+afHMXzgOTfHk3j9tof5z1+8Xsasz+Se6Z5K7T67y1qk7r7rdebmD3S5K8JEm2bdtmnxsAAIANLLPF761ZFL2dDkzyd0s87rwkR6yZPjzJ+bsaMx3Xd3CSz07z39Hdn+7uL2exlfHOS7wmAAAA6yxT/K7d3RftnJhuf9sSjzs1yVFVdYuqulaShyU5ed2Yk5M8erp9XJK3dXcnOSXJHavq26ZC+H1JPhQAAAD22DLF70tV9c2tbVV1lyRf2d2DpmP2nphFiftwkpO6+4yqelZVPXAadnySQ6pqR5KnZnE8Ybr7wiT/M4vyeHqS93b3G5Z/WwAAAOxUvZvTUVfVXbM4I+fO3TRvkuSh3X3azNn2yLZt23r79u1b82L72unE5b3y5J2XvPOSd14u5wDAXmY6L8q2je7b7clduvvUqvqOJLfN4mQsH+lu19QDAADYRyxzVs9kcXbNI6fx311V6e5XzJYKAACATbPb4ldVr0xyqyyOtdt5EfVOovgBAADsA5bZ4rctydG9u4MBAQAA2Cstc1bPDya58dxBAAAAmMcyW/wOTfKhqnpPkq/tnNndD9z1QwAAANhbLFP8njl3CAAAAOazzOUc3rEVQQAAAJjHbo/xq6p7VNWpVXVRVV1cVZdW1Re2IhwAAABX3TInd3lhkocn+ViSA5P81DQPAACAfcBSF3Dv7h1VtV93X5rkT6rqXTPnAgAAYJMsU/y+XFXXSnJ6Vf1ukk8muc68sQAAANgsy+zq+chp3BOTfCnJEUl+dM5QAAAAbJ5lit+Duvur3f2F7v7N7n5qkvvPHQwAAIDNsUzxe/QG8x6zyTkAAACYyS6P8auqhyd5RJJbVtXJa+66bpLPzB0MAACAzXFFJ3d5VxYncjk0yfPXzP9ikvfPGQoAAIDNs8vi192fqKrzknypu9+xhZkAAADYRFd4jN903b4vV9XBW5QHAACATbbMdfy+muQDVfW3WVzOIUnS3U+eLRUAAACbZpni94bpCwAAgH3Qbotfd59QVddKcptp1pnd/fV5YwEAALBZdlv8qupeSU5Ick6SSnJEVT26u/9h3mgAAABshmV29Xx+kvt095lJUlW3SfIXSe4yZzAAAAA2xxWe1XNyzZ2lL0m6+6NJrjlfJAAAADbTMlv8tlfV8UleOU3/WJLT5osEAADAZlqm+P1skickeXIWx/j9Q5IXzxkKAACAzbPMWT2/VlUvTPLWJN/I4qyeF8+eDAAAgE2xzFk9fyTJHyU5K4stfreoqsd395vmDgcAAMBVt+xZPb+/u3ckSVXdKosLuit+AAAA+4Blzur5qZ2lb3J2kk/NlAcAAIBNtswWvzOq6o1JTkrSSR6S5NSq+tEk6e6/mjEfAAAAV9Eyxe/aSf4jyfdN0xckuUGSB2RRBBU/AACAvdgyZ/X8ia0IAgAAwDyWOavnLZI8KcmRa8d39wPniwUAAMBmWWZXz9cmOT7J67K4jh8AAAD7kGWK31e7+w9mTwIAAMAslil+v19Vv5HkLUm+tnNmd793tlQAAABsmmWK3x2SPDLJvXPZrp49TQMAALCXW6b4PTjJLbv74rnDAAAAsPmuscSY9yW5/txBAAAAmMcyW/xulOQjVXVqLn+Mn8s5AAAA7AOWKX6/MXsKAAAAZrPb4tfd79iKIAAAAMxjl8Wvqr6Yxdk7v+WuJN3d15stFQAAAJtml8Wvu6+7lUEAAACYxzJn9QQAAGAfpvgBAAAMTvEDAAAYnOIHAAAwOMUPAABgcIofAADA4BQ/AACAwSl+AAAAg1P8AAAABqf4AQAADE7xAwAAGJziBwAAMDjFDwAAYHCKHwAAwOAUPwAAgMEpfgAAAINT/AAAAAan+AEAAAxO8QMAABic4gcAADA4xQ8AAGBwih8AAMDgFD8AAIDBKX4AAACDU/wAAAAGp/gBAAAMTvEDAAAYnOIHAAAwOMUPAABgcIofAADA4PZfdQAAGFbVqhNcpnvVCQBYIVv8AAAABqf4AQAADM6ungDAwt60a2pi91SATWSLHwAAwOAUPwAAgMEpfgAAAINT/AAAAAY3a/GrqvtW1ZlVtaOqnr7B/QdU1aum+99dVUeuu/8/VdVFVfW0OXMCAACMbLbiV1X7JXlRkvslOTrJw6vq6HXDHpvkwu6+dZIXJHnuuvtfkORNc2UEAAC4Ophzi9/dkuzo7rO7++IkJyY5dt2YY5OcMN1+TZJjqhbnkq6qByU5O8kZM2YEAAAY3pzF72ZJzl0zfd40b8Mx3X1Jks8nOaSqrpPkl5P85hW9QFU9rqq2V9X2Cy64YNOCAwAAjGTO4rfRVWDXX4l1V2N+M8kLuvuiK3qB7n5Jd2/r7m2HHXbYlYwJAAAwtv1nfO7zkhyxZvrwJOfvYsx5VbV/koOTfDbJ3ZMcV1W/m+T6Sb5RVV/t7hfOmBcAAGBIcxa/U5McVVW3SPJvSR6W5BHrxpyc5NFJ/jnJcUne1t2d5Ht2DqiqZya5SOkDAAC4cmYrft19SVU9MckpSfZL8rLuPqOqnpVke3efnOT4JK+sqh1ZbOl72Fx5AAAArq5qsYFt37dt27bevn371rxYbXRo4oos8/8n75Un77zknZe885J3foP8jgKwVarqtO7ettF9s17AHQAAgNVT/AAAAAan+AEAAAxO8QMAABic4gcAADA4xQ8AAGBwih8AAMDgFD8AAIDBKX4AAACDU/wAAAAGp/gBAAAMTvEDAAAYnOIHAAAwOMUPAABgcIofAADA4BQ/AACAwSl+AAAAg1P8AAAABqf4AQAADE7xAwAAGJziBwAAMDjFDwAAYHCKHwAAwOAUPwAAgMEpfgAAAINT/AAAAAan+AEAAAxO8QMAABic4gcAADA4xQ8AAGBwih8AAMDgFD8AAIDBKX4AAACDU/wAAAAGp/gBAAAMTvEDAAAYnOIHAAAwOMUPAABgcIofAADA4BQ/AACAwSl+AAAAg1P8AAAABqf4AQAADE7xAwAAGJziBwAAMDjFDwAAYHCKHwAAwOAUPwAAgMEpfgAAAINT/AAAAAan+AEAAAxO8QMAABic4gcAADA4xQ8AAGBwih8AAMDgFD8AAIDBKX4AAACDU/wAAAAGt/+qAwAAXClVq05wed2rTgCwS7b4AQAADE7xAwAAGJziBwAAMDjFDwAAYHCKHwAAwOAUPwAAgMEpfgAAAINT/AAAAAan+AEAAAxO8QMAABic4gcAADC4/VcdAADgaqFq1Qkur3vVCYAtZIsfAADA4BQ/AACAwSl+AAAAg1P8AAAABqf4AQAADE7xAwAAGJziBwAAMDjFDwAAYHAu4A4AwLdywXkYii1+AAAAg1P8AAAABqf4AQAADG7W4ldV962qM6tqR1U9fYP7D6iqV033v7uqjpzm/2BVnVZVH5j+vfecOQEAAEY2W/Grqv2SvCjJ/ZIcneThVXX0umGPTXJhd986yQuSPHea/+kkD+juOyR5dJJXzpUTAABgdHNu8btbkh3dfXZ3X5zkxCTHrhtzbJITptuvSXJMVVV3/9/uPn+af0aSa1fVATNmBQAAGNacxe9mSc5dM33eNG/DMd19SZLPJzlk3Zj/muT/dvfX1r9AVT2uqrZX1fYLLrhg04IDAACMZM7r+G108Zf1F2C5wjFVdbssdv+8z0Yv0N0vSfKSJNm2bZuLuwAAXF257iBcoTm3+J2X5Ig104cnOX9XY6pq/yQHJ/nsNH14kr9O8qjuPmvGnAAAAEObs/idmuSoqrpFVV0rycOSnLxuzMlZnLwlSY5L8rbu7qq6fpI3JPmV7n7njBkBAACGN1vxm47Ze2KSU5J8OMlJ3X1GVT2rqh44DTs+ySFVtSPJU5PsvOTDE5PcOsmvVdXp09cN58oKAAAwsupB9j/etm1bb9++fWtebG/ah3yZ/z95rzx55yXvvOSdl7zz211mea8aeec1yO/Y7Fuq6rTu3rbRfbNewB0AAIDVU/wAAAAGp/gBAAAMTvEDAAAY3JwXcAcAAHZlbzohjZPRDM8WPwAAgMEpfgAAAINT/AAAAAan+AEAAAxO8QMAABic4gcAADA4xQ8AAGBwih8AAMDgFD8AAIDBKX4AAACDU/wAAAAGp/gBAAAMTvEDAAAYnOIHAAAwOMUPAABgcIofAADA4BQ/AACAwSl+AAAAg1P8AAAABqf4AQAADE7xAwAAGJziBwAAMDjFDwAAYHCKHwAAwOAUPwAAgMEpfgAAAINT/AAAAAan+AEAAAxO8QMAABic4gcAADA4xQ8AAGBwih8AAMDgFD8AAIDBKX4AAACDU/wAAAAGp/gBAAAMTvEDAAAYnOIHAAAwOMUPAABgcIofAADA4BQ/AACAwSl+AAAAg1P8AAAABqf4AawEDL8AABA0SURBVAAADE7xAwAAGJziBwAAMDjFDwAAYHCKHwAAwOAUPwAAgMEpfgAAAINT/AAAAAan+AEAAAxO8QMAABic4gcAADA4xQ8AAGBwih8AAMDgFD8AAIDBKX4AAACDU/wAAAAGp/gBAAAMTvEDAAAYnOIHAAAwOMUPAABgcIofAADA4BQ/AACAwSl+AAAAg1P8AAAABqf4AQAADE7xAwAAGNz+qw4AAADsA6pWneAy3atOsM9R/AAAgPEoqpdjV08AAIDBKX4AAACDU/wAAAAGp/gBAAAMTvEDAAAYnOIHAAAwOMUPAABgcIofAADA4GYtflV136o6s6p2VNXTN7j/gKp61XT/u6vqyDX3/co0/8yq+qE5cwIAAIxstuJXVfsleVGS+yU5OsnDq+rodcMem+TC7r51khckee702KOTPCzJ7ZLcN8mLp+cDAABgD825xe9uSXZ099ndfXGSE5Mcu27MsUlOmG6/JskxVVXT/BO7+2vd/fEkO6bnAwAAYA/tP+Nz3yzJuWumz0ty912N6e5LqurzSQ6Z5v/LusfebP0LVNXjkjxumryoqs7cnOhb4tAkn77Kz1J11ZMsR975XfXM8l4Reecl77yufnkTPzN2Td55yTsveed1813dMWfx2+jd9ZJjlnlsuvslSV6y59FWr6q2d/e2VedYlrzz29cyyzsveecl77zknZe885J3XvKuzpy7ep6X5Ig104cnOX9XY6pq/yQHJ/nsko8FAABgCXMWv1OTHFVVt6iqa2VxspaT1405Ocmjp9vHJXlbd/c0/2HTWT9vkeSoJO+ZMSsAAMCwZtvVczpm74lJTkmyX5KXdfcZVfWsJNu7++Qkxyd5ZVXtyGJL38Omx55RVScl+VCSS5I8obsvnSvriuxru6jKO799LbO885J3XvLOS955yTsveecl74rUYgMbAAAAo5r1Au4AAACsnuIHAAAwOMVvRlV146o6sarOqqoPVdUbq+o2G4x7VFV9sKrOmMY9bS/P+7iq+sj09Z6quueK8h5eVX9TVR+bMv/+dCKhjcY+bcr7wap6X1U9agV5u6qevy7TM3cx9qCq+uPpfZ1RVf9QVeuvgzmrZZdvVb28qv6tqg6Ypg+tqnO2OOueLNuDq+oV03s6a7p98JaFvSzHssu3qupXp3Efraq/r6rbrSDvbj8fquqZVfXlqrrhmnkXrSDrnqwP51TVX66ZPq6qXj5/yg2z/Pfp+/39VXX6Rt/zVXWbadnvqKoPV9VJVXWjFeVdZp24bVW9fXo/H66qlR0rs7vlW1VHTuvOk9bMe2FVPWbLw2bp9eF2VfW26bPhY1X1a1Vbe8HGKceefM/9ZFV9YHpfH6yqY7cs6OVzLPsZfI+qeveadfiZK4i7M8vu1uFrVNUfTMv1A1V1ai1OkghJFL/ZTB+8f53k7d19q+4+Oskzktxo3bj7JXlKkvt09+2S3DnJ5/fivPdP8vgk9+zu70jyM0n+vKpuvIK8f5Xktd19VJLbJDkoybM3GPszSX4wyd26+/ZJvjcbXytybl9L8qNVdegSY/9PFic8OmpaLx6TxQVEt8SeLN/JpUl+covibWRPlu3xSc6e1vNbJfl4Fst7y+zh8n1Ckv+c5Lu6+zZJ/keSk6vq2lucd7efD5NPJ/mFrcq2C3uyPiTJtlWU6bWq6v9Lcv8kd+7uOyb5gSTnrhtz7SRvSPKH3X3r7v7OJH+Y5LAV5F12nfiDJC/o7jtNef/3FkdNstzynXwqyX/bqABspSXXhwOzOAv6c6bPhu/K4rPi57Y4brLk91xVHZ7kv2fxO8Qdk9wjyfu3IN/6HHvyGXxCksd1952S3D7JSVsWdI0l1+GHJrlpkjt29x2SPDjJ52bMtNQGgyvxvBetm35MVb3wqj7v9FyzZJ6e+8HTH0G+YzOebw6K33y+P8nXu/uPds7o7tO7+x/XjfuVJE/r7vOnMV/t7pduYc6dls37y0l+sbs/PY15bxYfik/YsqQL907y1e7+kynHpUl+PslPVtW3rRv7jCQ/191fmMZ+vrtP2NK0C5dkcWaon7+iQVV1qyR3T/Kr3f2NJOnus7v7DfNH/KY9Wb5J8r+S/Hwtrse5Cssu21snuUuS31oz+1lZ/OJ/q/nifYs9Wb6/nORJ3f3laexbkrwryY9tYd5lPx+S5GVJHlpVN9iydN9qqfVhjedl8TmxSjdJ8unu/lqSdPend/5cWOMRSf65u1+3c0Z3/313f3ALc+607DpxkyyuzbtzzAe2KN96yyzfJLkgyVtz2aWmVmXZ9eGd02dCps+IJyZ5+pYmXVj2e+6GSb6Y5KIk6e6LuvvjM2fbyJ58Bt8wySd3juvuD21p0ssss07cJMkn1/zucF53XzhHmD38g+BeYQsyPzzJP2W6SsHeSPGbz+2TnLaJ4+a2bI7bbTBu+zR/K31LjqnY/WuSW++cV1XXTXLd7j5ra+Pt0ouS/Fhd8a6Ft0ty+oovYbLU8l3jX7P4sHvk/NF2aZlle3TWLdvp9unZ2nV42fX3ekmus8H6u9Xfc3vyOXVRFuXvv80XZynLrA87nZTkztMfBlblLUmOmHbZe3FVfd8GY/aWnxfJ8llekORtVfWmqvr5qrr+zLl2ZZnlu9NzkvxCVe23Rdk2skzejT5Hzkpy0PTZsdWW+Z57X5L/SPLxqvqTqnrA1kT7FnvyM+4FSc6sqr+uqsdv5d4W6yyzTpyU5AHTbqDPr6rvnjHP0n8QrMXhK2+tqvdOu6CuZPfezJi5qg5K8l+SPDaKH4OrJFt9XZBdveb6+avItkvTD5ZXJHnyqrPsxrLLd63fSfKLWdHnypLL9sq8rzlc1Rx71Xq9gT9I8ugV/fKZZI+/1y5N8ntZ7IGxEt19URZbox+XxVanV9WKji3bTNMWle9M8uok90ryLzUdD7zFOZZevtMWqPdksUVtJZbMe0WfA1v++bDM99z0h7b7JjkuyUeTvGBFx8wt/Rnc3c9Ksi2L4vWIJG+ePd0Gllknuvu8JLfN4rPsG0neWlXHzBRpT/4Q9dUkD+7uO2dRvp4/bX3blQOn8np6VZ2exZ45m2HOzA9K8ubu/miSz1bVna9a1HkofvM5I4tv0MupqmevWZF3OW4Fls37oQ3G3Xmav5XOyOKD+JumXzKPSPLUKfMbpx9EX6qqW25xvivyv7L4i9B1kqSq9lvzAfesLN7bd1XVKr8/l1q+a+/v7h1ZbDn7/7cs5bdaZtl+99plO93+riQf3sKcV3X93ervuWU/H5Ik3f25JH+e1RxrtNbu1oe1XpnF8b//aYszftO0G9nbu/s3sthl76Fr8j4we8/Pi2QP1onuPr+7X9bdx2axS+DttzDnNy2xfNf6nSx2s17Z5/CS68P6z5FbJrmou7+4gsjJEt9zvfCe7v4fWWwZ+a8ryLlHP+O6+6zu/sMkx2Tx8/mQLU17WY7drsPd/bXuflN3/2IW6/GDVpF1nUryO1X1/iR/l+RmueLdK7/Si+OC79SLYyt/fStCrrOnmR+e5MTp9onT9N6nu33N8JXFCvPuJD+9Zt5dk3zfunE/nMVuWzeepg9I8uS9OO8Dk5ya5JBp+k5Z7BpxkxXk3Z7kUdP0fklemuT5G4z9uSRvSnK9afp6WRyovdXL+KI1t393Wm7P3MXYk7I4Dq2m6aOSHLuXLt+XJzluun27JOckOWcvXrZ/leTX10z/epK/3OK8e7J8n5zk9UkOnKZ/IMnZO6e3MO8ynw/PzOKY5WRxMqKPZ3EczZYt2yuxPpyT5NDp9s9NY1++gsy3zeJkTjunfzvJC9eNOTDJjiQ/smbefZPcYQV5l10n7pvkmtPtG2dxrNSN99Lle2SSD66ZPmlaHx6zl+Y9cPos+IE106/P4pjgrc671PdcFiceufOa6Z9K8voV5N2Tz+AfyWU/i78zixNY7beXrhN3TnLT6fY1stgC+7SZ8hyT5B+WHPuYJK9a81lwTpIjl1mf1jz+hVcm51ZkTnJIkq8k+cQ07tzpe6C2ej3Z3ZctfjPpxZrw4CQ/WNMp+bP4pej8dePemMV+8X83jTktyZafIGMP8p6cxfE776qqj2TxQfnj3f3JFeV9SFV9LItdRr6ajU/Q8IdJ/j7JqVX1wSTvSPLlrcq6C8/PFZ+l86ey+CVpR1V9IIvlvNGJCGaxh8t37ePOSPLe+RNeod0t28cmuU0tTod/VhZnc3vsliSb7OHy/d9Z/LHlA1V1ZpJfy+KPAF9ZQd4r/HxY95hPZ3EQ/Zbv1rfO7taHtY7PCj5/JwclOaEWZ5l7fxbHoz5z7YDp//z+SZ5Ui1PQfyiLX04+tcVZ92SduE+SD1bV+5KcksXJwf59S8Mu7Hb5buDZSQ6fO9guLLs+HJvkV6fPhg9k8VmxKWc/vAqu6HvumkmeV4vLK52exVkot/x44D38DH5kFsf4nZ7FngE/1qs5Bn+ZdfiGSV43/a7z/iy2sM+1PrwtyQFV9dM7Z1TVXWvjYw8PTvKp7v56VX1/kpvPlGl35sp8XJJXdPfNu/vI7j4iiz98ruRyZ1dk518wAAAAllJVN81i9967ZFGcz0nylO7+2LpxhyZ5XRbF//QsToJyv+4+ZxfPe1F3H7Rm+jFJtnX3E/fGzFX19iwuq/LmNfOenOQ7u/tnr2rmzaT4AQAADM6ungAAAINb1bEMAADAIKrqDlkcB7nW17r77huMPSTJWzd4mmO6+zNz5NvIvpj5qrCrJwAAwODs6gkAADA4xQ8AAGBwih8AV3tVdWlVnb7m68gr8RzXr6qf2/x0AHDVOcYPgKu99deNupLPcWSS13f37ffwcfut6ILQAFyN2OIHABuoqv2q6veq6tSqen9VPX6af1BVvbWq3ltVH6iqY6eHPCfJraYthr9XVfeqqteveb4XThciTlWdU1W/XlX/lOQhVXWrqnpzVZ1WVf9YVd+x1e8XgLG5nAMAJAdW1enT7Y9394OTPDbJ57v7rlV1QJJ3VtVbkpyb5MHd/YWqOjTJv1TVyUmenuT23X2nJKmqe+3mNb/a3fecxr41yc9098eq6u5JXpzk3pv9JgG4+lL8ACD5ys7CtsZ9ktyxqo6bpg9OclSS85L8TlV9b5JvJLlZkhtdidd8VbLYgpjkPyd5dVXtvO+AK/F8ALBLih8AbKySPKm7T7nczMXumocluUt3f72qzkly7Q0ef0kuf0jF+jFfmv69RpLPbVA8AWDTOMYPADZ2SpKfraprJklV3aaqrpPFlr9PTaXv+5PcfBr//9q7Q5wGYDAKwO8pFOEeGAw32AFQqEmOwwlIsBjOskAgg+AIQWPgAEVsArEsQWDK9+n+eal8aZt+JTn8Mf+W5LjtQdujJItdIWOMzySvbc+3OW178jdbAuC/UvwAYLfrJM9J7tquk1xlc1PmJslp21WSZZKXJBljfGTzDnDd9nKM8Z7kNsnjduZ+T9YyyUXbhyRPSc72rAWAX/OdAwAAwOSc+AEAAExO8QMAAJic4gcAADA5xQ8AAGByih8AAMDkFD8AAIDJKX4AAACT+waFMXN5FNhrigAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 1080x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "'''\n",
    "features based on combination of:\n",
    "atom_types = ['C','N','O,'S']\n",
    "'''\n",
    "\n",
    "atom_types = ['C','N','O','S']\n",
    "feature_labels = []\n",
    "for i in range(len(atom_types)):\n",
    "    for j in range(len(atom_types)):\n",
    "        label = atom_types[i]+'-'+atom_types[j]\n",
    "        feature_labels.append(label)\n",
    "feature_labels+=['C_a_H', 'C_a_A']\n",
    "feature_labels=np.array(feature_labels)\n",
    "\n",
    "importance = rf.feature_importances_\n",
    "\n",
    "#indices = np.argsort(importance)[::-1][:20]\n",
    "indices = np.argsort(importance)[::-1]\n",
    "\n",
    "print(\"Top Feature ranking:\")\n",
    "n = importance.shape[0]\n",
    "#n = 20\n",
    "for f in range(n):\n",
    "    print(\"%d. feature %d (%f)\" % (f + 1, indices[f], importance[indices[f]]))\n",
    "\n",
    "\n",
    "plt.figure()\n",
    "plt.title(\"Feature importances\")\n",
    "plt.bar(range(n), importance[indices],\n",
    "       color=\"r\", align=\"center\")\n",
    "plt.xticks(range(n), feature_labels[indices])\n",
    "plt.xlim([-1, n])\n",
    "plt.xlabel('Feature')\n",
    "plt.ylabel('Importance value')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 8 Angstrom cutoff"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training Features Shape: (1812, 18)\n",
      "Training Labels Shape: (1812,)\n",
      "Testing Features Shape: (604, 18)\n",
      "Testing Labels Shape: (604,)\n"
     ]
    }
   ],
   "source": [
    "data_file = os.getcwd()+'/Data/dataset_ha_beta_8A.pkl'\n",
    "dataset = dataset_loader(data_file)\n",
    "\n",
    "features = np.array([data['x_vector'] for data in dataset])\n",
    "labels = np.array([data['y'] for data in dataset])\n",
    "x_train, x_test, y_train, y_test = train_test_split(features, labels, test_size=0.25, random_state=13)\n",
    "print('Training Features Shape:', x_train.shape)\n",
    "print('Training Labels Shape:', y_train.shape)\n",
    "print('Testing Features Shape:', x_test.shape)\n",
    "print('Testing Labels Shape:', y_test.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\beryl\\Anaconda3\\lib\\site-packages\\sklearn\\base.py:318: UserWarning: Trying to unpickle estimator DecisionTreeRegressor from version 0.21.2 when using version 0.22.1. This might lead to breaking code or invalid results. Use at your own risk.\n",
      "  UserWarning)\n",
      "C:\\Users\\beryl\\Anaconda3\\lib\\site-packages\\sklearn\\base.py:318: UserWarning: Trying to unpickle estimator RandomForestRegressor from version 0.21.2 when using version 0.22.1. This might lead to breaking code or invalid results. Use at your own risk.\n",
      "  UserWarning)\n"
     ]
    }
   ],
   "source": [
    "'''Model Loader'''\n",
    "model_file= os.getcwd()+\"/Model/rf_pp_ha_beta_8A_split.pkl\"\n",
    "with open(model_file, \"rb\") as f:\n",
    "    rf = pickle.load(f)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=12)]: Using backend ThreadingBackend with 12 concurrent workers.\n",
      "[Parallel(n_jobs=12)]: Done  26 tasks      | elapsed:    0.0s\n",
      "[Parallel(n_jobs=12)]: Done 176 tasks      | elapsed:    0.0s\n",
      "[Parallel(n_jobs=12)]: Done 426 tasks      | elapsed:    0.0s\n",
      "[Parallel(n_jobs=12)]: Done 776 tasks      | elapsed:    0.0s\n",
      "[Parallel(n_jobs=12)]: Done 1000 out of 1000 | elapsed:    0.0s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean Absolute Error: 1.39\n",
      "Accuracy: 77.99 %.\n",
      "Root Mean Squared Error : 1.81\n",
      "Pearson Correlation Coefficient : 0.42\n"
     ]
    }
   ],
   "source": [
    "'''Model evaluation on Test set'''\n",
    "#Mean Absolute Error\n",
    "preds = rf.predict(x_test)\n",
    "errors = abs(preds - y_test)\n",
    "print('Mean Absolute Error:', round(np.mean(errors), 2))\n",
    "\n",
    "#Mean Absolute Percentage Error & Accuracy\n",
    "mape = 100 * (errors / y_test)\n",
    "accuracy = 100 - np.mean(mape)\n",
    "print('Accuracy:', round(accuracy, 2), '%.')\n",
    "\n",
    "#Root Mean Squared Error\n",
    "rmse = sqrt(mean_squared_error(y_test, preds))\n",
    "print('Root Mean Squared Error :', round(rmse, 2))\n",
    "\n",
    "#Pearson Correlation Coefficient (PCC) score\n",
    "pcc = pearsonr(y_test, preds)\n",
    "print('Pearson Correlation Coefficient :', round(pcc[0],2))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "\n",
    "fig_size = np.zeros(2)\n",
    "fig_size[0] = 15\n",
    "fig_size[1] = 10\n",
    "plt.rcParams[\"figure.figsize\"] = fig_size\n",
    "\n",
    "f, ax = plt.subplots()\n",
    "\n",
    "plt.plot(y_test, preds, 'o', markerfacecolor=\"None\", markeredgecolor='green', markeredgewidth=1)\n",
    "#ax.plot(ax.get_xlim(), ax.get_ylim(), ls=\"--\", c=\".3\")\n",
    "coef = np.polyfit(y_test,preds,1)\n",
    "poly1d_fn = np.poly1d(coef)\n",
    "plt.plot(y_test, poly1d_fn(y_test), '--k')\n",
    "\n",
    "plt.title(\"Test data binding affinity\")\n",
    "plt.xlabel('Actual binding affinity data')\n",
    "plt.ylabel('Predicted binding affinity data')\n",
    "#plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Top Feature ranking:\n",
      "1. feature 16 (0.080201)\n",
      "2. feature 8 (0.073643)\n",
      "3. feature 0 (0.069616)\n",
      "4. feature 17 (0.067297)\n",
      "5. feature 9 (0.063912)\n",
      "6. feature 4 (0.063104)\n",
      "7. feature 2 (0.059922)\n",
      "8. feature 6 (0.058220)\n",
      "9. feature 10 (0.057627)\n",
      "10. feature 14 (0.057096)\n",
      "11. feature 1 (0.054647)\n",
      "12. feature 12 (0.053135)\n",
      "13. feature 5 (0.051490)\n",
      "14. feature 13 (0.050674)\n",
      "15. feature 3 (0.044765)\n",
      "16. feature 11 (0.037767)\n",
      "17. feature 7 (0.037497)\n",
      "18. feature 15 (0.019388)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "'''\n",
    "features based on combination of:\n",
    "atom_types = ['C','N','O,'S']\n",
    "'''\n",
    "\n",
    "atom_types = ['C','N','O','S']\n",
    "feature_labels = []\n",
    "for i in range(len(atom_types)):\n",
    "    for j in range(len(atom_types)):\n",
    "        label = atom_types[i]+'-'+atom_types[j]\n",
    "        feature_labels.append(label)\n",
    "feature_labels+=['C_a_H', 'C_a_A']\n",
    "feature_labels=np.array(feature_labels)\n",
    "\n",
    "importance = rf.feature_importances_\n",
    "\n",
    "#indices = np.argsort(importance)[::-1][:20]\n",
    "indices = np.argsort(importance)[::-1]\n",
    "\n",
    "print(\"Top Feature ranking:\")\n",
    "n = importance.shape[0]\n",
    "#n = 20\n",
    "for f in range(n):\n",
    "    print(\"%d. feature %d (%f)\" % (f + 1, indices[f], importance[indices[f]]))\n",
    "\n",
    "\n",
    "plt.figure()\n",
    "plt.title(\"Feature importances\")\n",
    "plt.bar(range(n), importance[indices],\n",
    "       color=\"r\", align=\"center\")\n",
    "plt.xticks(range(n), feature_labels[indices])\n",
    "plt.xlim([-1, n])\n",
    "plt.xlabel('Feature')\n",
    "plt.ylabel('Importance value')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### 16 Angstrom cutoff"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training Features Shape: (1812, 18)\n",
      "Training Labels Shape: (1812,)\n",
      "Testing Features Shape: (604, 18)\n",
      "Testing Labels Shape: (604,)\n"
     ]
    }
   ],
   "source": [
    "data_file = os.getcwd()+'/Data/dataset_ha_beta_16A.pkl'\n",
    "dataset = dataset_loader(data_file)\n",
    "\n",
    "features = np.array([data['x_vector'] for data in dataset])\n",
    "labels = np.array([data['y'] for data in dataset])\n",
    "x_train, x_test, y_train, y_test = train_test_split(features, labels, test_size=0.25, random_state=13)\n",
    "print('Training Features Shape:', x_train.shape)\n",
    "print('Training Labels Shape:', y_train.shape)\n",
    "print('Testing Features Shape:', x_test.shape)\n",
    "print('Testing Labels Shape:', y_test.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "'''Model Loader'''\n",
    "model_file= os.getcwd()+\"/Model/rf_pp_ha_beta_16A_split.pkl\"\n",
    "with open(model_file, \"rb\") as f:\n",
    "    rf = pickle.load(f)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=12)]: Using backend ThreadingBackend with 12 concurrent workers.\n",
      "[Parallel(n_jobs=12)]: Done  26 tasks      | elapsed:    0.0s\n",
      "[Parallel(n_jobs=12)]: Done 176 tasks      | elapsed:    0.0s\n",
      "[Parallel(n_jobs=12)]: Done 426 tasks      | elapsed:    0.0s\n",
      "[Parallel(n_jobs=12)]: Done 776 tasks      | elapsed:    0.0s\n",
      "[Parallel(n_jobs=12)]: Done 1000 out of 1000 | elapsed:    0.0s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean Absolute Error: 1.37\n",
      "Accuracy: 78.16 %.\n",
      "Root Mean Squared Error : 1.76\n",
      "Pearson Correlation Coefficient : 0.46\n"
     ]
    }
   ],
   "source": [
    "'''Model evaluation on Test set'''\n",
    "#Mean Absolute Error\n",
    "preds = rf.predict(x_test)\n",
    "errors = abs(preds - y_test)\n",
    "print('Mean Absolute Error:', round(np.mean(errors), 2))\n",
    "\n",
    "#Mean Absolute Percentage Error & Accuracy\n",
    "mape = 100 * (errors / y_test)\n",
    "accuracy = 100 - np.mean(mape)\n",
    "print('Accuracy:', round(accuracy, 2), '%.')\n",
    "\n",
    "#Root Mean Squared Error\n",
    "rmse = sqrt(mean_squared_error(y_test, preds))\n",
    "print('Root Mean Squared Error :', round(rmse, 2))\n",
    "\n",
    "#Pearson Correlation Coefficient (PCC) score\n",
    "pcc = pearsonr(y_test, preds)\n",
    "print('Pearson Correlation Coefficient :', round(pcc[0],2))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "\n",
    "fig_size = np.zeros(2)\n",
    "fig_size[0] = 15\n",
    "fig_size[1] = 10\n",
    "plt.rcParams[\"figure.figsize\"] = fig_size\n",
    "\n",
    "f, ax = plt.subplots()\n",
    "\n",
    "plt.plot(y_test, preds, 'o', markerfacecolor=\"None\", markeredgecolor='green', markeredgewidth=1)\n",
    "#ax.plot(ax.get_xlim(), ax.get_ylim(), ls=\"--\", c=\".3\")\n",
    "coef = np.polyfit(y_test,preds,1)\n",
    "poly1d_fn = np.poly1d(coef)\n",
    "plt.plot(y_test, poly1d_fn(y_test), '--k')\n",
    "\n",
    "plt.title(\"Test data binding affinity\")\n",
    "plt.xlabel('Actual binding affinity data')\n",
    "plt.ylabel('Predicted binding affinity data')\n",
    "#plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Top Feature ranking:\n",
      "1. feature 17 (0.133003)\n",
      "2. feature 16 (0.092227)\n",
      "3. feature 10 (0.075436)\n",
      "4. feature 9 (0.056581)\n",
      "5. feature 14 (0.050453)\n",
      "6. feature 2 (0.050185)\n",
      "7. feature 11 (0.049568)\n",
      "8. feature 12 (0.048493)\n",
      "9. feature 15 (0.047658)\n",
      "10. feature 0 (0.047536)\n",
      "11. feature 1 (0.047291)\n",
      "12. feature 8 (0.045693)\n",
      "13. feature 5 (0.045661)\n",
      "14. feature 13 (0.045379)\n",
      "15. feature 6 (0.043925)\n",
      "16. feature 3 (0.040830)\n",
      "17. feature 7 (0.040780)\n",
      "18. feature 4 (0.039300)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "'''\n",
    "features based on combination of:\n",
    "atom_types = ['C','N','O,'S']\n",
    "'''\n",
    "\n",
    "atom_types = ['C','N','O','S']\n",
    "feature_labels = []\n",
    "for i in range(len(atom_types)):\n",
    "    for j in range(len(atom_types)):\n",
    "        label = atom_types[i]+'-'+atom_types[j]\n",
    "        feature_labels.append(label)\n",
    "feature_labels+=['C_a_H', 'C_a_A']\n",
    "feature_labels=np.array(feature_labels)\n",
    "\n",
    "importance = rf.feature_importances_\n",
    "\n",
    "#indices = np.argsort(importance)[::-1][:20]\n",
    "indices = np.argsort(importance)[::-1]\n",
    "\n",
    "print(\"Top Feature ranking:\")\n",
    "n = importance.shape[0]\n",
    "#n = 20\n",
    "for f in range(n):\n",
    "    print(\"%d. feature %d (%f)\" % (f + 1, indices[f], importance[indices[f]]))\n",
    "\n",
    "\n",
    "plt.figure()\n",
    "plt.title(\"Feature importances\")\n",
    "plt.bar(range(n), importance[indices],\n",
    "       color=\"r\", align=\"center\")\n",
    "plt.xticks(range(n), feature_labels[indices])\n",
    "plt.xlim([-1, n])\n",
    "plt.xlabel('Feature')\n",
    "plt.ylabel('Importance value')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=12)]: Using backend ThreadingBackend with 12 concurrent workers.\n",
      "[Parallel(n_jobs=12)]: Done  26 tasks      | elapsed:    0.0s\n",
      "[Parallel(n_jobs=12)]: Done 176 tasks      | elapsed:    0.0s\n",
      "[Parallel(n_jobs=12)]: Done 426 tasks      | elapsed:    0.0s\n",
      "[Parallel(n_jobs=12)]: Done 776 tasks      | elapsed:    0.0s\n",
      "[Parallel(n_jobs=12)]: Done 1000 out of 1000 | elapsed:    0.0s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.02534273 0.01981853 0.01173228 0.00944142 0.00585488 0.00804027\n",
      " 0.00968839 0.00587544 0.00647286 0.04778359 0.0448579  0.03421148\n",
      " 0.01340655 0.01570419 0.02341884 0.01592081 0.1742104  0.2088501 ]\n",
      "[17 16  9 10 11  0 14  1 15 13 12  2  6  3  5  8  7  4]\n"
     ]
    }
   ],
   "source": [
    "'''\n",
    "permutation importance on test data\n",
    "'''\n",
    "result = permutation_importance(rf, x_test, y_test, n_repeats=10,\n",
    "                                random_state=42, n_jobs=-1)\n",
    "\n",
    "sorted_idx = result.importances_mean.argsort()[::-1]\n",
    "p_importance_mean_test = result.importances_mean\n",
    "print(p_importance_mean_test)\n",
    "print(sorted_idx)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# data to plot\n",
    "length = len(sorted_idx)\n",
    "\n",
    "atom_types = ['C','N','O','S']\n",
    "feature_labels = []\n",
    "for i in range(len(atom_types)):\n",
    "    for j in range(len(atom_types)):\n",
    "        label = atom_types[i]+'-'+atom_types[j]\n",
    "        feature_labels.append(label)\n",
    "feature_labels+=['C_a_H', 'C_a_A']\n",
    "feature_labels=np.array(feature_labels)\n",
    "indices = sorted_idx\n",
    "# create plot\n",
    "\n",
    "plt.figure()\n",
    "plt.title(\"Permutation feature importances\")\n",
    "plt.bar(range(length), p_importance_mean_test[indices],\n",
    "       color=\"g\", align=\"center\")\n",
    "plt.xticks(range(length), feature_labels[indices])\n",
    "plt.xlim([-1, length])\n",
    "plt.xlabel('Feature')\n",
    "plt.ylabel('Importance value')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Comparison of R and feature importance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 193,
   "metadata": {},
   "outputs": [],
   "source": [
    "'''data values'''\n",
    "cutoffs = np.array([4,8,12,16])\n",
    "PCC_R = np.array([0.35, 0.42, 0.43, 0.46])\n",
    "rmses = np.array([1.87, 1.81, 1.8, 1.76])\n",
    "feature_importances = np.zeros((4, 18))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 194,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\beryl\\Anaconda3\\lib\\site-packages\\sklearn\\base.py:318: UserWarning: Trying to unpickle estimator DecisionTreeRegressor from version 0.21.2 when using version 0.22.1. This might lead to breaking code or invalid results. Use at your own risk.\n",
      "  UserWarning)\n",
      "C:\\Users\\beryl\\Anaconda3\\lib\\site-packages\\sklearn\\base.py:318: UserWarning: Trying to unpickle estimator RandomForestRegressor from version 0.21.2 when using version 0.22.1. This might lead to breaking code or invalid results. Use at your own risk.\n",
      "  UserWarning)\n"
     ]
    }
   ],
   "source": [
    "'''model loader to get the feature importance values'''\n",
    "rfs = []\n",
    "for cutoff in cutoffs:\n",
    "    with open(os.getcwd()+\"/Model/rf_pp_ha_beta_\"+str(cutoff)+\"A_split.pkl\", \"rb\") as f:\n",
    "        rfs.append(pickle.load(f))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 195,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[0.11175806 0.07763539 0.10346004 0.04170897 0.09623754 0.0603526\n",
      "  0.08304709 0.02164767 0.10210291 0.08890647 0.07782158 0.01987789\n",
      "  0.04236383 0.02611647 0.02358577 0.01799616 0.00287621 0.00250534]\n",
      " [0.06961555 0.05464723 0.05992185 0.04476529 0.06310407 0.05148989\n",
      "  0.05822023 0.03749711 0.07364296 0.06391224 0.05762686 0.03776707\n",
      "  0.05313479 0.05067361 0.05709618 0.01938788 0.08020055 0.06729665]\n",
      " [0.05662473 0.04310486 0.04960245 0.04103954 0.04982183 0.04319387\n",
      "  0.0498449  0.04300576 0.06010076 0.05972551 0.07552061 0.04024279\n",
      "  0.04765858 0.05356842 0.04251898 0.03396121 0.10170944 0.10875577]\n",
      " [0.04753566 0.04729072 0.0501853  0.04083004 0.0393004  0.04566095\n",
      "  0.04392532 0.04078007 0.04569337 0.05658063 0.07543642 0.04956777\n",
      "  0.04849293 0.04537927 0.05045304 0.04765827 0.09222696 0.13300286]]\n"
     ]
    }
   ],
   "source": [
    "'''fill the feature_importances'''\n",
    "for idx, rf in enumerate(rfs):\n",
    "    feature_importances[idx]=rf.feature_importances_\n",
    "print(feature_importances)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 196,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAcAAAAFNCAYAAACXC791AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjAsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+17YcXAAAgAElEQVR4nO3dd3xV9f3H8deHsCFhSBDCChsZMoxgHXVbB67WhdiqtbX+FFfdVlsctXXbKmqtVVsVcVQtAs66rYOwIewhhLBX2GR8fn+cE7yEmxAgNze59/18PPLgjO8553MH+eR7Pud8j7k7IiIiyaZWvAMQERGJByVAERFJSkqAIiKSlJQARUQkKSkBiohIUlICFBGRpKQEKNWWmb1gZvfGO44SZnagmX1uZhvN7OEqOuYxZpZbFcfaG2bWwMzeMbMNZvZ6uOxeM1ttZsvjHZ9IRdSOdwBSs5jZIuBAoAjYBLwHDHP3TfGMq4pcDqwG0jzBbqA1sxeAXHe/o4KbnEPwPTjA3QvNrB1wA9DB3VfGKMy4MLPhQBd3vyjesUjlUg9Q9sXp7t4Y6Af0B26LczxVpQOQk2jJbx91AOa4e2HE/JqqSH5mVqP+cK9p8SYTJUDZZ+6+HHifIBFGZWYzzWxwxHzt8DTZgHD+dTNbHp5K+9zMepWxn0vM7MtSy9zMuoTT9czsITNbbGYrzOxpM2sQrmthZmPMbL2ZrTWzL8ws6nffzA43s/FhPOPN7PBw+QvAxcDNZrbJzE6Ism15MTQLY1hlZuvC6bYR2zY3s+fNLC9c/3apfd9gZivNbJmZXVrO+x11P+W9f2Z2OTA04rW9E64/yMw+Dd+3GWZ2Rrj8LuD3wPlh+98AHwIZ4fwLUeI6xsxyzez28PNfZGZDI9afZmaTzCzfzJaEva6SdZlhrJeZ2WLg43B5md+d8PT5k2b2bhjTV2bWysweC9+XWWbWP6J9hpn9O/x8FprZNeHyk4HbI17rlHB5EzP7R/h5LA1P/6ZEvNdfmdmjZrYWGB6+z5+Fsa42s1fL+gyl6igByj4Lf4GfAswrp9krwJCI+Z8Aq919Yjj/LtAVaAlMBF7ex3DuB7oRJOMuQBuCX9IQnJrLBdIJTtvdDuzWizOz5sBY4K/AAcAjwFgzO8DdLwlje8DdG7v7R3sZQy3geYKeUntgK/BExLYvAg2BXgTvxaMR61oBTcL9XQaMMLNmZbwP5e0nKnd/ptRrO93M6gDvAB+E+7kaeNnMurv7H4D7gFfD9n8j+B7khfOXlHGoVkCL8HVcDDxjZt3DdZuBXwBNgdOA/zOzs0ptfzRwEMF3CPb83TkPuCM85nbg67BdC+ANgs+X8I+hd4ApYWzHA9eZ2U/c/b1Sr7VvuO9/AoUEn3N/4CTgVxHHHgQsCGP7I3BP+F42A9oCj5fxHklVcnf96KfCP8AigtrfRoIk8l+gaTntu4RtG4bzLwO/L6Nt03CfTcL5F4B7w+lLgC9Ltfdw/0bwC7RzxLofAQvD6buB/xDUccp7bT8Hviu17GvgktLxRNm23BiitO8HrAunWwPFQLMo7Y4hSJa1I5atBA6L0ra8/ZT5/kV7bcBRwHKgVsSyV4Dh4fRw4KVSceaW894eQ5AwGkUsew24s4z2jwGPhtOZYaydytl/tO/O3yPWXw3MjJjvA6wPpwcBi0vt7zbg+TJe64EECbVBxLIhwCcR73Xp/f0LeAZoWxn/D/VTOT/qAcq+OMvdUwl+qfUg+IsaM2sfnibaZGabANx9HjATON3MGgJnACPD9ilm9mczm29m+QTJlZL97YV0gl7PhPB03XqCi3PSw/UPEvRSPzCzBWZ2axn7yQC+L7Xse4JewX7FYGYNzexvZvZ9+Fo/B5qGp83aAWvdfV0Z+17jP9TaALYAjaO029N+9kYGsMTdiyOWVfS9KMs6d99can8ZAGY2yMw+CU9BbgCuYPfvwZKSiQp+d1ZETG+NMl/yHnYgOH27PuKzu50g0UXTAagDLIto/zeC3t5usYZuJvgj6bvwdPIvy9i3VCElQNln7v4ZwV/aD4Xziz04TdTYg4tkSpScBj2T4CKSklOmF4bLTiA4xZcZLrcoh9tMkGCCBmatItatJviF1svdm4Y/TUpicPeN7n6Du3cCTgd+a2bHRzlGHsEvt0jtgaXlvA0VioHgNGx3YJC7pwE/jnitS4DmZta0AscpT3n7Ke/9g91PCecB7WzXWmlF34uyNDOzRqX2lxdOjwRGA+3cvQnwNLt/DyJj3Jvvzp4sIeipN434SXX3U6Mct6T9dqBFRPs0d4+sX++yjbsvd/dfu3sG8BvgSQvr1xI/SoCyvx4DTjSzMi+EAUYR1Ej+j7D3F0ol+EWyhuCX833l7GMK0MvM+plZfYLTUgCEvZS/A4+aWUsAM2tjZj8JpweHFyEYkE9wC0dRlGOMA7qZ2YUWXKxzPtATGFNOXBWKIXytW4H1Ya3xDxHbLiOoZz1pwcUydczsx+ylPeynzPcvtALoFDH/LUHSvDnczzEEfzyM2tu4SrnLzOqa2VHAYOD1cHkqQe91m5kNJEhw5dmb786efAfkm9ktFtzfmGJmvc3s0HD9CiCz5I+B8H3+AHjYzNLMrJaZdTazo8s6gJmdaz9c9LSOIEFG+w5KFVIClP3i7qsI6ht3ltNmGUEt7XAg8uq3fxGcBlsK5ADflLOPOQS1vI+AucCXpZrcQnCa85vwlNhHBD0uCC6U+Iigdvk18KS7fxrlGGsIfinfQPCL9WZgsLuvLiuuvYjhMaABQU/xG4LTo5F+DhQAswhqfNdV8JilRd1PBd6/fwA9w1N6b7v7DoLT1aeEMT8J/MLdZ+1jXBDUFNcR9PpeBq6I2N+VwN1mtpHgwqHX9rCvCn939sTdiwiSez9gIcHrfZagZwk/JOk1ZlZy8dYvgLrhsdcRXFTTupzDHAp8G5YGRgPXuvvCfY1ZKoe565YmEYmtsAf5kru33VNbkaqiHqCIiCQlJUAREUlKOgUqIiJJST1AERFJSkqAIiKSlBJmlPIWLVp4ZmZmvMMQEZFqZMKECavdPT3auoRJgJmZmWRnZ8c7DBERqUbMrPTwhjvpFKiIiCQlJUAREal+iov33GY/KQGKiEj14g7DhgX/xpASoIiIVC9jx8JTT8G4cTE9jBKgiIhUH+4wfHgwPXx4THuBSoAiIlJ9jB0LEyYE09nZMe0FKgGKiEj1ENn7KxHDXmBME6CZnWxms81snpndWk67c8zMzSwrYtnBZva1mc0ws2nhQzxFRCRRRfb+SsSwFxizBGhmKcAIggdq9gSGmFnPKO1SgWsInkBdsqw28BLBAzN7AccQPORTREQSUbTeX4kY9QJj2QMcCMxz9wXh06VHAWdGaXcP8ACwLWLZScBUd58CwZO6w6c2i4hIIorW+ysRo15gLBNgG2BJxHxuuGwnM+sPtHP3MaW27Qa4mb1vZhPN7OZoBzCzy80s28yyV61aVZmxi4hIVSmv91ciBr3AWCZAi7JsZ/RmVgt4FLghSrvawJHA0PDfs83s+N125v6Mu2e5e1Z6etSxTkVEpJorHP3OLr2/WYf8mFX52+DUU39oFINeYCwTYC7QLmK+LZAXMZ8K9AY+NbNFwGHA6PBCmFzgM3df7e5bgHHAgBjGKiIicfDJrBXMH3bTLst6TPic9H89u3vCq+ReYCwT4Higq5l1NLO6wAXA6JKV7r7B3Vu4e6a7ZwLfAGe4ezbwPnCwmTUML4g5GsiJYawiIlKF5q7YyMXPfceLdzxJ99w5uzcYNmz3ZZXcC4zZ45DcvdDMhhEksxTgOXefYWZ3A9nuPrqcbdeZ2SMESdSBce4+NlaxiohI1Vi3eQePfTSHl75dTMM6tfh4+pt7t4Phw4NToxatyrZ3zGM82GhVycrKcj0PUESketpRWMyL33zPXz6aw6bthQwd1IGbCuaSdt5P935n77wDgwdXqKmZTXD3rGjrEuaBuCIiUv24Ox/PWskfx85kwerNHNW1BXec1pPurcLLQOLYCVMCFBGRmJi1PJ97x8zky3mr6ZTeiOcvOZRjuqdjlXD6sjIoAYqISKVas2k7j3w4h1e+W0xq/Tr84fSeXHRYB+qkVK/hp5UARUSkUmwvLOKf/1vE4/+dx5aCIn7xo0yuO6ErTRvWjXdoUSkBiojIfnF3PshZwX3jZvL9mi0c2z2d3512EF1apsY7tHIpAYqIyD7LycvnnjE5fL1gDV1bNuafvxzI0d1qxshcSoAiIrLXVm3czsMfzObV7CU0bVCHe87sxZCB7aldzep85VECFBGRCttWUMRzXy3kyU/ms62giMuO6MjVx3WlScM68Q5trykBiojIHrk7705fzn3jZpK7bisnHHQgt5/ag07pjeMd2j5TAhQRkXJNy93APWNy+G7RWnq0SuXlXw3iiC4t4h3WflMCFBGRqFbkb+PB92fz74m5NG9Yl/vO7sP5h7YjpVb1uJF9fykBiojILrYVFPHsFwt48tP5FBQVc/lRnbjquC6k1a95db7yKAGKiAgQ1PnembqM+9+dxdL1Wzm5VytuO7UHHQ5oFO/QYkIJUEREmLxkPfeMyWHC9+vo2TqNh87ty486HxDvsGJKCVBEJIkt27CVB9+bzZuTltKicT3u/1kfzjkkcep85VECFBFJQlt3FPG3z+fz9GfzKXa48pjOXHlsFxrXS560kDyvVEREKC52Rk/J4/73ZrFswzZO69OaW0/pQbvmDeMdWpVTAhQRSRITvl/HPWNymLxkPX3aNOEvF/RnYMfm8Q4rbpQARUQS3NL1W7n/3VmMnpJHy9R6PHRuX37avw21kqDOVx4lQBGRBLV5eyFPfzafZz5fAMDVx3XhiqM70yiJ6nzl0bsgIpJgioudNyct5YH3ZrFy43bO6JvBLaf0oE3TBvEOrVpRAhQRSSDjF63l7ndymLZ0A33bNeWpiw7hkA7N4h1WtaQEKCKSAJas3cKf353F2GnLaJVWn8fO78cZfTOSvs5XHiVAEZEabNP2Qp78ZB7PfrmQFDOuO6Erl/+4Ew3r6tf7nugdEhGpgYqKnTcmLOHB9+ewetN2ftq/DTed3J3WTVTnqyglQBGRGubr+Wu4Z0wOOcvyGdC+Kc9enEW/dk3jHVaNowQoIlJDfL9mM/eNm8n7M1bQpmkDHh/Sn8EHt8ZMdb59oQQoIlLN5W8rYMTH83j+q0XUTjFuPKkbvzqqE/XrpMQ7tBpNCVBEpJoqKnZGjV/MIx/MYe2WHZwzoC03/qQ7B6bVj3doCUEJUESkGvpq3mruGZPDrOUbGZjZnBcG96RP2ybxDiuhKAGKiFQjC1dv5o9jZ/LRzBW0bdaAJ4cO4JTerVTniwElQBGRamDDlgL++vFc/vX1IurVTuGWk3tw6RGZqvPFkBKgiEgcFRYV88p3i3nkwzms31rA+Vnt+O1J3WiZqjpfrCkBiojEyWdzVnHvmBzmrtzEYZ2ac+fgnvTKUJ2vqigBiohUsXkrN/HHsTl8MnsVHQ5oyN9+fggn9TxQdb4qViuWOzezk81stpnNM7Nby2l3jpm5mWWVWt7ezDaZ2Y2xjFNEpCqs27yD4aNn8JPHPid70Tp+d+pBfHD9j/lJL13kEg8x6wGaWQowAjgRyAXGm9lod88p1S4VuAb4NspuHgXejVWMIiJVoaComJe++Z7HPprLxm0FDBnYnutP7EaLxvXiHVpSi+Up0IHAPHdfAGBmo4AzgZxS7e4BHgB26eWZ2VnAAmBzDGMUEYkZd+eT2Su5d+xMFqzazJFdWnDH4IPo0Sot3qEJsU2AbYAlEfO5wKDIBmbWH2jn7mMiT3OaWSPgFoLeo05/ikiNM2fFRu4Zk8MXc1fTqUUj/nFxFsf1aKlTndVILBNgtE/Zd640q0VwivOSKO3uAh51903lfVnM7HLgcoD27dvvT6wiIpVi7eYdPPrhHEZ+t5hGdVO4c3BPfn5YB+rWjuklF7IPYpkAc4F2EfNtgbyI+VSgN/BpmORaAaPN7AyCnuI5ZvYA0BQoNrNt7v5E5AHc/RngGYCsrCxHRCROdhQW86+vF/GX/85ly44iLhrUnutO6EazRnXjHZqUIZYJcDzQ1cw6AkuBC4ALS1a6+wagRcm8mX0K3Oju2cBREcuHA5tKJz8RkerA3flo5kr+ODaHRWu2cHS3dO447SC6Hpga79BkD2KWAN290MyGAe8DKcBz7j7DzO4Gst19dKyOLSJSFWYuy+fesTl8NW8NXVo25vlLD+XY7i3jHZZUkLknxpnDrKwsz87OjncYIpIEVm/azsMfzOHV8YtJa1CH60/oxoWD2lMnRXW+6sbMJrh7VrR1GglGRKSCthcW8cJXi3j843lsKyji4sMzufb4rjRtqDpfTaQEKCKyB+7O+zOWc9+4WSxeu4Xje7Tk9tMOonN643iHJvtBCVBEpBzTl27gnjE5fLtwLd0ObMyLlw3kqK7p8Q5LKoESoIhIFCs3buOh92fz+oRcmjWsy71n9eaCQ9tRW3W+hKEEKCISYVtBEf/4ciFPfjKPHUXF/OrIjgw7ritNGtSJd2hSyZQARUQI6nzjpi3nvnEzWbp+Kyf1PJDbTz2IzBaN4h2axIgSoIgkvam567lnTA7jF62jR6tURv5qEId3abHnDaVGUwIUkaS1fMM2Hnh/Fm9OXEqLxnX580/7cG5WO1JqacDqZKAEKCJJZ+uOIv7+xQKe+nQ+RcXOFUd35qpjO5NaX3W+ZKIEKCI1W3Ex1KrYlZnuzugpedz/7izyNmzj1D6tuPXkg2h/QMMYBynVkRKgiNRc7jBsGIwYAXt4zt7Exeu4Z0wOkxavp1dGGo+e349BnQ6ookClOlICFJGaa+xYeOopOO204CeKvPVbeeC9Wbw9OY/01Ho8eM7B/GxAW2qpzpf0lABFpGZyh+HDg+nhw+HUU3fpBW7ZUcjTny3gmc/nU+ww7NguXHFMZxrX0689CeibICI109ixMGFCMJ2dDePGwWmnUVzsvD15Kfe/N4sV+dsZfHBrbj2lB22bqc4nu1ICFJGaJ7L3V2L4cCb0Ooy7x8xkSu4G+rZtwogLB5CV2TwuIUr1pwQoIjVPZO+vRHY2T9z0OMv7H8kj5/XlrH5tVOeTcikBikjNEq33F/rT1H+T9tLvaFhP9/PJnmlYcxGpEbYVFDE1dz2fPfrC7r2/UKs502n40QdVG5jUWOoBiki1s37LDnLy8slZls+MvHxy8vKZt2oTRUXFjP7XfeVvHOWKUJFolABFJG7cnbwN25ixdMMuyW7p+q0727RKq0/PjDRO6nUgR8/5loOXz/thB6edBmPGBP+OGxcsi7giVKQ8SoAiUiUKi4pZsHozM/I2kJMXJrtl+azfUgAEHbZOLRoxoEMzfv6jDvRsnUbPjDRaNK4X7MAdbh+6607Hjg1GgSlJfiXUC5QKUAIUkUq3ZUchs5Zv3Nmjy8nbwKzlG9leWAxAvdq16NEqlVN6t6JnRhN6ZaTRo1UqDeuW8ysp2pWfEAyFVpp6gVIBSoAisl/WbNq+8/TljDDZLVy9mWIP1jdpUIdeGWn84kcd6JmRRq+MJnRq0YjaKXtxDV45V36WSb1A2QMlQBGpEHdnydqtwSnMiHrd8vxtO9u0adqAnhlpnN43g56t0+jVpgkZTepj+5uEyur9lSc7O9hu8OD9O7YkLCVAEdnNjsJi5q3ctEuym5mXz8bthQCk1DK6pDfmR50PoFdG2s56XdOGdWMT0ODBQS9QpBIpAYokuY3bCoJ6XcSVmHNXbGJHUVCva1AnhYNap3Jm/wx6ZTShZ+s0urdKpX6dlDhHLrJ/lABFksjK/G3MWFZyYUo+M/I2sGjNlp3rD2hUl54ZaVx6ZCa9wotTMg9oRIqGFJMEpAQokoCKi51FazaXujgln9Wbtu9s0755Q3plpPGzAW3p1Sa4OKVlar39r9eJ1BBKgCI13PbCIuYs30TOsg07k93MZfls2VEEQO1aRtcDUzmme/rOet1BGWmk1dd4mZLclABFapANWwt2nrrMCU9lzlu5icLwnoPG9WrTs3Ua52W1o2eY7Loe2Jh6tVWvEylNCVCkGnJ3ludvY8bSkhFTgt5d7rofhghrmVqPXhlpHH9Qy50Xp7Rv3lCPABKpICVASQ7FxVCrej78pKjYWbBq0y731s3I28C6iCHCOh7QiH7tmnLhoPY7k116ar04Ry5SsykBSuJzD4bLGjEi7qOCbN1RxKzluz7lYNbyfLYVBLcc1K1di+4HpvKTXq2Cel1GGj1apdGonv6rilQ2/a+SxDd2LDz1VDAuZBWODblu845dTl/OyMtnwapNO4cIS6tfm54ZaQwd1GFnsuuc3pg6ezNEmIjsMyVASWyRY0jGaGxIdyd33dadTzfIyQsS3rINPwwRltGkPj0zmnBqn9Y7r8Rs26yBbjkQiSMlQElskWNIVsITAgqKipm/atMuF6fk5OWTvy0YIqyWQef0xgzq2HznwM8HtU6jeaMYDREmIvsspgnQzE4G/gKkAM+6+5/LaHcO8DpwqLtnm9mJwJ+BusAO4CZ3/ziWsUoCivYEgb3oBW7eXsis5eGN5EuD3t3sFRvZET7Sp36dWvRoFQ78HCa7HhoiTKTGiFkCNLMUYARwIpALjDez0e6eU6pdKnAN8G3E4tXA6e6eZ2a9gfeBNrGKVRJUtCcIlNELXLVx+24DPy9cs3nn+MvNGtahV0YTLj08M0x2aXRs0VhDhInUYLHsAQ4E5rn7AgAzGwWcCeSUancP8ABwY8kCd58UsX4GUN/M6rn7dkQqopznx22/4/d81LY/Oct/uBJz5cYfvlrtmjegZ+s0zurfZufFKa3SKuGRPiJSrcQyAbYBlkTM5wKDIhuYWX+gnbuPMbMbie5nwCQlP9kr5Tw/rt7kibxx7zN80XUgXVo25siuLXYO/HxQ6zSaNNAQYSLJIJYJMNqfyzsf6GVmtYBHgUvK3IFZL+B+4KQy1l8OXA7Qvn37/QhVEkoFnh4+Yu5oar18B/Xr6jowkWQVyxuOcoF2EfNtgbyI+VSgN/CpmS0CDgNGm1kWgJm1Bd4CfuHu86MdwN2fcfcsd89KT0+PwUuQGqlU7++/nQ/lpa8XBRe/hBpOmUT9D9+v+thEpNqIZQIcD3Q1s45mVhe4ABhdstLdN7h7C3fPdPdM4BvgjPAq0KbAWOA2d/8qhjFKoonS+zt+/niGZr8TXPwSafhwPWVcJInFLAG6eyEwjOAKzpnAa+4+w8zuNrMz9rD5MKALcKeZTQ5/WsYqVkkgZdT+7Oqrd29bckWoiCQl8wT5CzgrK8uzs7PjHYbEkzscemiZF79ElZUF330X9zFCRSQ2zGyCu2dFW6dBByVxlHPlZ5mys4PtRCTp6BI4SRyDB/O/eau46NlvOaNvBo+e30/37olImdQDlISxauN2rh01mcwWjfjj2X2U/ESkXOoBSkIoKnaue3USG7cV8OJlA/X8PBHZI/2WkITw+Mdz+WreGh742cH0aJUW73BEpAbQKVCp8b6at5q//HcuPx3QhnOz2sY7HBGpIZQApUZbuXEb146aTOf0xtx7Vm/V/USkwnQKVGqsomLn2lcms3l7ISN/PYiGGtdTRPaCfmNIjfWXj+bw9YI1PHRuX7odmBrvcESkhtEpUKmRPp+zisc/mcc5h7TlnENU9xORvacEKDXOivxtXP/qZLq2bMw9Z/aOdzgiUkMpAUqNUlhUzNWvTGJrQRFPDh1Ag7op8Q5JRGqoCidAM+tgZieE0w3MTEUXqXKPfjSH7xau5Y9n96ZLS30FRWTfVSgBmtmvgTeAv4WL2gJvxyookWg+m7OKEZ/M5/ysdpzdX3U/Edk/Fe0BXgUcAeQDuPtcQM/nkyqzbMNWrn91Mj1apXLXmb3iHY6IJICKJsDt7r6jZMbMagOJ8SBBqfYKi4q55pVJbCsoYsTQAdSvo7qfiOy/iibAz8zsdqCBmZ0IvA68E7uwRH7w0AdzGL9oHX/6aR86pzeOdzgikiAqmgBvBVYB04DfAOOAO2IVlEiJT2at5OnP5jNkYHvO7Ncm3uGISAKp0Egw7l4M/D38EakSeeu3cv1rkzmodRp/OL1nvMMRkQRToQRoZguJUvNz906VHpEIUFBUzLCREykoLOZJ1f1EJAYqOhZoVsR0feBcoHnlhyMSeOj92UxcvJ7Hh/SnY4tG8Q5HRBJQhWqA7r4m4mepuz8GHBfj2CRJ/XfmCv72+QIuOqw9p/fNiHc4IpKgKnoKdEDEbC2CHqGG4ZBKl7tuC799bQq9MtK44zTV/UQkdip6CvThiOlCYBFwXqVHI0ltR2Exw0ZOoqjYGXGh6n4iElsVvQr02FgHIvLAe7OYvGQ9Tw4dQKbqfiISY+UmQDP7bXnr3f2Ryg1HktUHM5bz7JcLufhHHTi1T+t4hyMiSWBPPUDV+STmlqzdwo2vT6FPmybcftpB8Q5HRJJEuQnQ3e+qqkAkOQV1v4k4MOLCAdSrrbqfiFSNil4FWh+4DOhFcB8gAO7+yxjFJUniT+/OZEruBp6+6BDaH9Aw3uGISBKp6FigLwKtgJ8AnxE8D3BjrIKS5PDe9GU8/9UiLj0ik5N7t4p3OCKSZCqaALu4+53AZnf/J3Aa0Cd2YUmiW7xmCze9MZW+bZtw2ymq+4lI1atoAiwI/11vZr2BJkBmTCKShLe9sIirRk7EgCcuHEDd2hX9GoqIVJ6K3gj/jJk1A+4ERgONw2mRvXbf2JlMW7qBZ35+CO2aq+4nIvFR0QT4vLsXEdT/9AQI2Wdjpy7jn19/z2VHduSkXqr7iUj8VPTc00Ize8bMjjczi2lEkrAWrd7MLf+eSr92Tbnl5B7xDkdEklxFE2B34CPgKmCRmT1hZkfGLixJNNsKgrpfSi3jiQv7q+4nInFX0cchbXX319z9p0A/II3gdGi5zOxkM5ttZvPM7NZy2p1jZm5mWRHLbgu3m21mP6lInFJ93Ts2hxl5+Tx8bl/aNlPdT0Tir8J/hpvZ0Wb2JDCR4Gb4cp8GYWYpwAjgFKAnMMTMdnu+jZmlAtcA30Ys6wlcQHDj/cnAk+H+pAZ6Z0oeL32zmMt/3IkTeh4Y704I2DYAABgJSURBVHBERIAKJkAzWwhcB3wB9Hb389z933vYbCAwz90XuPsOYBRwZpR29wAPANsilp0JjHL37e6+EJgX7k9qmIWrN3Pbm9M4pEMzbvpJ93iHIyKyU0V7gH3d/Wx3f8XdN1dwmzbAkoj53HDZTmbWH2jn7mP2dlup/rYVFHHlyxOpnWI8PqQ/dVJU9xOR6qOiNcD8kmkzm1jBfUe7WtQj9lMLeBS4YW+3jdjH5WaWbWbZq1atqmBYUlXuHpPDzGX5PHpePzKaNoh3OCIiu9iXP8krehtELtAuYr4tkBcxnwr0Bj41s0XAYcDo8EKYPW0LgLs/4+5Z7p6Vnp5e8VcgMfefyUsZ+e1irji6M8f2aBnvcEREdrMvCXBsBduNB7qaWUczq0twUcvokpXuvsHdW7h7prtnAt8AZ7h7dtjuAjOrZ2Ydga7Ad/sQq8TB/FWbuP3NaWR1aMaNJ3WLdzgiIlHtdQJ09zvMLMXMhu6hXSEwDHgfmAm85u4zzOxuMztjD9vOAF4DcoD3gKvCkWikmtu6o4irXp5IvTopPH5hf2qr7ici1ZS571Za+2GlWRrBze9tCHplH4bzNwGT3T3aVZ1xkZWV5dnZ2fEOI+nd+u+pjBq/hBcuPZRjuuvUp4jEl5lNcPesaOv2NBboi8A64GvgVwSJry5wprtPrtQopcZ7a1Iuo8Yv4apjOyv5iUi1t6cE2Mnd+wCY2bPAaqC9u+thuLKLeSs3cvub0xnYsTnXn6C6n4hUf3sq0JQ8B5CwBrdQyU9K27ojuN+vYd0UHh+iup+I1Ax76gH2NbOSewANaBDOG+DunhbT6KRG+P1/pjN35Sb+9cuBHJhWP97hiIhUSLkJ0N01/qaU640Jubw+IZdrjuvCUV11L6aI1Bw6VyX7bM6Kjdzx9jQO69Sca1X3E5EaRglQ9smWHYVc+fJEGterw18v6E9KLT0nWURqlj3VAEV24+7c8fZ05q/axEuXDaKl6n4iUgOpByh77fXsXN6cuJRrjuvKEV1axDscEZF9ogQoe2XW8nzu/M90juhyANcc3zXe4YiI7DMlQKmwzdsLuerliaQ1qMNj56vuJyI1mxKgVIi787u3prFw9Wb+ckE/0lPrxTskEZH9ogQoFTJq/BLenpzHdSd04/DOqvuJSM2nBCh7lJOXzx9Gz+Cori246tgu8Q5HRKRSKAFKuTZtL2TYyIk0bVCHR8/vp7qfiCQM3QcoZXJ3bntzGovWbGbkrw+jRWPV/UQkcagHKGUa+d1i3pmSxw0ndeewTgfEOxwRkUqlBChRTV+6gbveyeHobun839Gd4x2OiEilUwKU3WzcVsCwkRNp3rAuj5zXl1qq+4lIAlINUHbh7tz65jSWrNvKqMsP4wDV/UQkQakHKLt46ZvvGTt1GTee1J1DM5vHOxwRkZhRApSdpuVu4J4xMzm2ezq/+XGneIcjIhJTSoACQP62Aq4aOZEDGtfl4fP6qe4nIglPNUDB3bnljankrd/Kq785jOaN6sY7JBGRmFMPUPjn/xbx7vTl3Hxydw7poLqfiCQHJcAkN2XJev44bibH92jJr45U3U9EkocSYBLbsDWo+7VMrc/Dut9PRJKMaoBJyt25+Y0pLN+wjdeu+BFNG6ruJyLJRT3AJPXcV4t4f8YKbj2lBwPaN4t3OCIiVU4JMAlNWryOP42byYk9D+SyIzvGOxwRkbhQAkwy67fsYNjISbRqUp+HzumLmep+IpKcVANMIu7Oja9PZeXGbbx+xeE0aVgn3iGJiMSNeoBJ5NkvFvLRzBXcdspB9GvXNN7hiIjElRJgkpjw/Truf28WJ/dqxaVHZMY7HBGRuFMCTALrNu/g6pETad20Pvefc7DqfiIiqAaY8IqLnRten8LqTTt44/9+RJMGqvuJiECMe4BmdrKZzTazeWZ2a5T1V5jZNDObbGZfmlnPcHkdM/tnuG6mmd0WyzgT2TNfLODjWSv53WkHcXBb1f1ERErELAGaWQowAjgF6AkMKUlwEUa6ex937wc8ADwSLj8XqOfufYBDgN+YWWasYk1U2YvW8uD7szmtT2t+8aMO8Q5HRKRaiWUPcCAwz90XuPsOYBRwZmQDd8+PmG0EeMkqoJGZ1QYaADuAyLayB2s3B/f7tW3WgD/9rI/qfiIipcSyBtgGWBIxnwsMKt3IzK4CfgvUBY4LF79BkCyXAQ2B6919bQxjTSjFxc71r05m7eYdvHnl4aTVV91PRKS0WPYAo3U5fLcF7iPcvTNwC3BHuHggUARkAB2BG8xst2f1mNnlZpZtZtmrVq2qvMhruKc/n89nc1Zx5+k96d2mSbzDERGplmKZAHOBdhHzbYG8ctqPAs4Kpy8E3nP3AndfCXwFZJXewN2fcfcsd89KT0+vpLBrtu8WruXhD+Yw+ODWXDSofbzDERGptmKZAMcDXc2so5nVBS4ARkc2MLOuEbOnAXPD6cXAcRZoBBwGzIphrAlh9abtXP3KRNo3b8iffqq6n4hIeWJWA3T3QjMbBrwPpADPufsMM7sbyHb30cAwMzsBKADWAReHm48AngemE5xKfd7dp8Yq1kRQUvdbt6WA5y45lFTV/UREyhXTG+HdfRwwrtSy30dMX1vGdpsIboWQCnry03l8MXc1953dh14ZqvuJiOyJhkJLAF/PX8MjH87hzH4ZDBnYbs8biIiIEmBNt2rjdq4ZNYnMAxrxx7NV9xMRqSiNBVqDFYV1v/ytBfzrlwNpXE8fp4hIRek3Zg32xMfz+HLeau7/WR8Oap0W73BERGoUnQKtof43bzWP/XcOZ/dvw3lZqvuJiOwtJcAaaOXGbVwzajKdWjTi3rN6q+4nIrIPdAq0hikqdq59ZTKbthfw8q8G0Uh1PxGRfaLfnjXMX/47l68XrOGBcw6me6vUeIcjIlJj6RRoDfLF3FU8/vFcfjagrep+IiL7SQmwhliRv43rRk2mS3pj7jmrV7zDERGp8XQKtAYoLCrmmlcmsWVHEaMuH0DDuvrYRET2l36T1gCPfTSXbxeu5eFz+9L1QNX9REQqg06BVnOfzVnFiE/ncV5WW352SNt4hyMikjCUAKux5Ru2cf2rk+nWMpW7zugd73BERBKKEmA1VVL321ZQxIihA2hQNyXeIYmIJBTVAKuphz+cw3eL1vLY+f3o0rJxvMMREUk46gFWQ5/MXslTn85nyMB2nNW/TbzDERFJSEqA1Uze+q389tXJ9GiVyh9O1/1+IiKxogRYjRQUFXP1K5PYUVjMk0MHUL+O6n4iIrGiGmA18tAHs5nw/Tr+OqQ/ndJV9xMRiSX1AKuJ/85cwd8+W8DQQe05o29GvMMREUl4SoDVwNL1W7nh9Sn0bJ3GnYN7xjscEZGkoAQYZwVFxQwbOZHCIlfdT0SkCqkGGGcPvDeLSYvX88SF/cls0Sje4YiIJA31AOPow5wV/P2Lhfz8sA4MPlh1PxGRqqQEGCdL1m7hhtcm07tNGncMPije4YiIJB0lwDjYUVjMsFcm4Q4jLhxAvdqq+4mIVDXVAOPgz+/OYsqS9Tw1dAAdDlDdT0QkHtQDrGLvTV/Oc18t5JLDMzmlT+t4hyMikrSUAKvQkrVbuOmNKfRt24TbTu0R73BERJKaEmAV2V5YxFUjJwLwhOp+IiJxpxpgFfnTuFlMzd3A0xcdQrvmDeMdjohI0lMPsAqMm7aMF/63iF8e0ZGTe7eKdzgiIoISYMx9v2Yzt7wxlb7tmnLrKar7iYhUF0qAMbStIKj7mcETQ/pTt7bebhGR6kI1wBj649iZTF+az99/kaW6n4hINRPTLomZnWxms81snpndGmX9FWY2zcwmm9mXZtYzYt3BZva1mc0I29SPZayVbczUPF785nt+fVRHTux5YLzDERGRUmKWAM0sBRgBnAL0BIZEJrjQSHfv4+79gAeAR8JtawMvAVe4ey/gGKAgVrFWtoWrN3Prv6fRv31Tbj5ZdT8Rkeoolj3AgcA8d1/g7juAUcCZkQ3cPT9ithHg4fRJwFR3nxK2W+PuRTGMtdJsKyjiqpcnUjvFeOLCAdRJUd1PRKQ6iuVv5zbAkoj53HDZLszsKjObT9ADvCZc3A1wM3vfzCaa2c0xjLNS3T0mh5xl+TxyXl/aNG0Q73BERKQMsUyAFmWZ77bAfYS7dwZuAe4IF9cGjgSGhv+ebWbH73YAs8vNLNvMsletWlV5ke+j/0xeyshvF/OboztxXA/V/UREqrNYJsBcoF3EfFsgr5z2o4CzIrb9zN1Xu/sWYBwwoPQG7v6Mu2e5e1Z6enolhb1v5q/axO1vTuOQDs248aTucY1FRET2LJYJcDzQ1cw6mlld4AJgdGQDM+saMXsaMDecfh842MwahhfEHA3kxDDW/VJS96tbuxZPXNhfdT8RkRogZvcBunuhmQ0jSGYpwHPuPsPM7gay3X00MMzMTiC4wnMdcHG47Toze4QgiTowzt3HxirW/XXXOzOYtXwjz196KK2bqO4nIlITxPRGeHcfR3D6MnLZ7yOmry1n25cIboWo1t6etJRXvlvClcd05tjuLeMdjoiIVJDO1e2HeSs3cftb0xiY2Zzfntgt3uGIiMheUALcR1t3BHW/BnVS+OuQ/tRW3U9EpEbRWKD76A+jpzNn5Ub+eelAWjWpUaO0iYgI6gHuk39PyOW17FyuOqYLP+4W39svRERk3ygB7qW5KzZyx9vTGdSxOded0HXPG4iISLWkBLgXtuwo5MqXJ9KoXgqPq+4nIlKjqQa4F+58ewbzVm3ixV8OomWa6n4iIjWZujAV9Fr2Ev49MZerj+vKkV1bxDscERHZT0qApRUX77Zo9vKN/P4/0zm88wFce7zqfiIiiUAJMJI7DBsW/BvavL2QK1+eQON6dXjsgn6k1Ir2kAsREalplAAjjR0LTz0F44LR29ydO96ezsLVm/nrBf1omaq6n4hIolACLOEOw4cH08OHgzuvjl/CW5OWcu3x3Ti8i+p+IiKJRFeBlhg7FiZMCKazs1ny0hv8YXZjjuzSgmHHdYlvbCIiUunUA4Rde3+hzbffSVr92jx6vup+IiKJSAkQdu39hXrkzubF1qtJT60Xp6BERCSWlACj9P5K9Hj6kV2uCBURkcShBBil97dTdvbOK0JFRCSxJHcCLKf3t1N4RaiIiCSW5E6ApXp/BSefwoyl6+HUU39oo16giEhCSt4EGKX3V+e9d+n11ku7Jzz1AkVEEk7yJsCyan/Dhu2+TL1AEZGEk5wJsCK1v9LUCxQRSSjJmQDLu/KzLNnZwXYiIpIQknMotMGD1ZsTEUlyydkDFBGRpKcEKCIiSUkJUEREkpISoIiIJCUlQBERSUpKgCIikpTME+R2ADNbBXxfCbtqAayuhP1I9aLPNXHps01MlfW5dnD39GgrEiYBVhYzy3b3rHjHIZVLn2vi0mebmKric9UpUBERSUpKgCIikpSUAHf3TLwDkJjQ55q49Nkmpph/rqoBiohIUlIPUEREkpISYAQzSzGzSWY2Jt6xSOUxs+vNbIaZTTezV8ysfrxjkn1jZs+Z2Uozmx6x7EEzm2VmU83sLTNrGs8YZe9F+1zD5Veb2ezw/+8DlX1cJcBdXQvMjHcQUnnMrA1wDZDl7r2BFOCC+EYl++EF4ORSyz4Eerv7wcAc4LaqDkr22wuU+lzN7FjgTOBgd+8FPFTZB1UCDJlZW+A04Nl4xyKVrjbQwMxqAw2BvDjHI/vI3T8H1pZa9oG7F4az3wBtqzww2S/RPlfg/4A/u/v2sM3Kyj6uEuAPHgNuBorjHYhUHndfSvCX42JgGbDB3T+Ib1QSQ78E3o13EFIpugFHmdm3ZvaZmR1a2QdQAgTMbDCw0t0nxDsWqVxm1ozgNEpHIANoZGYXxTcqiQUz+x1QCLwc71ikUtQGmgGHATcBr5mZVeYBlAADRwBnmNkiYBRwnJm9FN+QpJKcACx091XuXgC8CRwe55ikkpnZxcBgYKjr3q5EkQu86YHvCM7OtajMAygBAu5+m7u3dfdMggskPnZ39RISw2LgMDNrGP71eDy60CmhmNnJwC3AGe6+Jd7xSKV5GzgOwMy6AXWp5EHPlQAlobn7t8AbwERgGsF3XiOH1FBm9grwNdDdzHLN7DLgCSAV+NDMJpvZ03ENUvZaGZ/rc0Cn8NaIUcDFld2710gwIiKSlNQDFBGRpKQEKCIiSUkJUEREkpISoIiIJCUlQBERSUpKgJIwzKyVmY0ys/lmlmNm48L7h8rb5joza1iBfZ9rZjPN7JNw/pXw6QPX72G7TeG/GWb2RjntmprZlXuKI1Ys8LGZpUUsO9vM3Mx6xOB4mWZ2YWXvN2L/w8zs0ljtXxKDEqAkhPAm97eAT929s7v3BG4HDtzDptcRDJC9J5cBV7r7sWbWCjjc3Q9290crEp+757n7OeU0aQrELQECpwJT3D0/YtkQ4Eti8/SMTCBqAgwHLd9fzxE8BUSkTEqAkiiOBQrcfedN0O4+2d2/MLNjIp/xaGZPmNklZnYNwfign0T07IaY2bTw2YH3h8t+DxwJPG1mDwIfAC3Dm66PigzCzDqa2ddmNt7M7olYnlnyrDMz62Vm34XbTzWzrsCfgc7hsgfNrLGZ/dfMJobxnBmxn5lm9vfwGWkfmFmDcF0XM/vIzKaE23UOl98UxjPVzO4q4/0bCvwnIt7GBEMEXkZEAgzfy0/N7A0LnsH3csn4jGZ2arjsSzP7a8l7bmZHh69rsgXP20wNX+9R4bLrw8/jdTN7B/gg7JE+GH4O08zs/Ijjf2Zmr5nZHDP7s5kNDd/PaSWvORwRZpGZDdzTF0eSmLvrRz81/ofgr/1Hy1h3DDAmYv4J4JJwehHQIpzOIBg6LZ1gIN6PgbPCdZ8SPFMQgt7L9DKONRr4RTh9FbCp9DbA4wRjVkIwvFOD0vsMj58WTrcA5gEWtisE+oXrXgMuCqe/Bc4Op+sT9GxPIhj5xgj+4B0D/DhK3N8DqRHzFwH/CKf/BwyIeC83EDxyqBbB6B1HhsdbAnQM271S8p4D7wBHhNONw9dW+jO5hGDsx+bh/M8InvOXQtCLXwy0DrdbH07XA5YCd4XbXAs8FrHP3wE3xPu7qZ/q+6MeoMgPDiU4hbrKg+fLvQz8eC/3cQTBL3+AF8to8zVwu5ndAnRw961R2hhwn5lNBT4C2vDD6dyF7j45nJ4AZIa9qjbu/haAu2/zoBd0UvgziWA4uB5A1yjHa+7uGyPmhxAMP0X475CIdd+5e667FwOTCZJyD2CBuy8M27wS0f4r4JGwx93Uf3h2X2kfunvJM+GOBF5x9yJ3XwF8RvD5AIx392UePCduPkGPHIKh7jIj9reS4I8akagq41y7SHUwAyirxlbIrqf765fRrrIetVLu+ILuPtLMviV4APP7ZvYrYEGpZkMJeqKHuHuBBU8qKYl7e0S7IoIeZFmxG/And//bHmIuNLNa7l5sZgcQDELc28ycoBfmZnZzGcevXc7xcfc/m9lYgjrjN2Z2QhlNN5eKuyyRxy+OmC9m199p9YFof1yIAKoBSuL4GKhnZr8uWWBmh5rZ0QSn93qaWT0za0LwRIgSGwkGUobgFOLRZtbCzFIIej2f7WUcX/FDzWxotAZm1omgt/RXglOmB5eKA6AJwTMqC8zsWKBDeQf14OKVXDM7KzxGPQuubn0f+GVY08PM2phZyyi7mA10CqfPAf7l7h3cPdPd2wELCXplZZlFMHBxZjh/fsTr7ezu09z9fiCboLdY+vWW9jlwvpmlmFk6QU/8u3LaR9MNmL6X20gSUQKUhODuDpwNnGjBbRAzgOFAnrsvIaiVTSU4rTkpYtNngHfN7BN3XwbcBnwCTAEmuvt/2DvXAleZ2XiCJBbN+cB0M5tMkAz+5e5rgK/Ciz4eDOPMMrNsgkQ6qwLH/jlwTXja9H9AK3f/ABgJfG1m0wiejBEt8YwlqK9BkPjfKrX+35Rx1SZAeBr3SuA9M/sSWEFQKwS4LnxdUwh6ZO8SfBaF4QU70W4leStsM4Xgj5ub3X15eS8+iiMITh+LRKWnQYgIZtaaIBGfuB/7aOzum8KrQkcAc72Ct4lUNjPrD/zW3X8ej+NLzaAeoIgQ9n7/bhE3wu+DX4e92hkEvd891R1jqQVwZxyPLzWAeoAiIpKU1AMUEZGkpAQoIiJJSQlQRESSkhKgiIgkJSVAERFJSkqAIiKSlP4fv36xjIQ9M18AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 504x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "'''R comparison'''\n",
    "fig_size = np.zeros(2)\n",
    "fig_size[0] = 7\n",
    "fig_size[1] = 5\n",
    "plt.rcParams[\"figure.figsize\"] = fig_size\n",
    "plt.plot(cutoffs, PCC_R, '^-', markerfacecolor=\"None\", markeredgecolor='red', markeredgewidth=5)\n",
    "plt.title(\"R-values of each cutoff parameters\")\n",
    "plt.xlabel('Cutoff distance (Angstrom)')\n",
    "plt.ylabel('R-value')\n",
    "plt.xticks(cutoffs)\n",
    "#plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 212,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 504x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "'''R and rmse'''\n",
    "fig_size = np.zeros(2)\n",
    "fig_size[0] = 7\n",
    "fig_size[1] = 5\n",
    "plt.rcParams[\"figure.figsize\"] = fig_size\n",
    "\n",
    "t = cutoffs\n",
    "data1 = PCC_R\n",
    "data2 = rmses\n",
    "\n",
    "'''\n",
    "fig, ax1 = plt.subplots()\n",
    "color = 'tab:red'\n",
    "ax1.set_xlabel('cutoff')\n",
    "ax1.set_ylabel('R-value', color=color)\n",
    "ax1.plot(t, data1, '^-', color=color, markeredgewidth=5)\n",
    "ax1.tick_params(axis='y', labelcolor=color)\n",
    "ax1.set_xticks(cutoffs)\n",
    "ax1.legend('R-value', loc='center left')\n",
    "\n",
    "ax2 = ax1.twinx()  # instantiate a second axes that shares the same x-axis\n",
    "\n",
    "color = 'tab:blue'\n",
    "ax2.set_ylabel('RMSE', color=color)  # we already handled the x-label with ax1\n",
    "ax2.plot(t, data2, 'o-',color=color, markeredgewidth=5)\n",
    "ax2.tick_params(axis='y', labelcolor=color)\n",
    "ax2.legend('RMSE', loc='center right')\n",
    "fig.tight_layout()  # otherwise the right y-label is slightly clipped\n",
    "plt.show()\n",
    "\n",
    "'''\n",
    "fig, ax = plt.subplots()\n",
    "\n",
    "\n",
    "l1, l2 = ax.plot(t, data1, '^-', t, data2, 'o-')\n",
    "\n",
    "ax.legend((l1, l2), ('oscillatory', 'damped'), loc='upper right', shadow=True)\n",
    "ax.set_xlabel('time')\n",
    "ax.set_ylabel('volts')\n",
    "ax.set_title('Damped oscillation')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 152,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 936x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "'''feature importance comparisons'''\n",
    "# data to plot\n",
    "n_groups = feature_importances[0].shape[0]\n",
    "\n",
    "# create plot\n",
    "fig_size = np.zeros(2)\n",
    "fig_size[0] = 13\n",
    "fig_size[1] = 10\n",
    "plt.rcParams[\"figure.figsize\"] = fig_size\n",
    "\n",
    "fig, ax = plt.subplots()\n",
    "index = np.arange(n_groups)\n",
    "bar_width = 0.8\n",
    "opacity = 0.8\n",
    "\n",
    "rects1 = plt.bar(index - bar_width/4, feature_importances[0], bar_width/4,\n",
    "alpha=opacity,\n",
    "color='r', hatch='-',\n",
    "label='4 Angstrom')\n",
    "\n",
    "rects2 = plt.bar(index, feature_importances[1], bar_width/4,\n",
    "alpha=opacity,\n",
    "color='g',hatch='x',\n",
    "label='8 Angstrom')\n",
    "\n",
    "rects2 = plt.bar(index + bar_width/4, feature_importances[2], bar_width/4,\n",
    "alpha=opacity,\n",
    "color='b',hatch='/',\n",
    "label='12 Angstrom')\n",
    "\n",
    "rects2 = plt.bar(index + 2*bar_width/4, feature_importances[3], bar_width/4,\n",
    "alpha=opacity,\n",
    "color='black',hatch='o',\n",
    "label='16 Angstrom')\n",
    "\n",
    "plt.xlabel('Feature by index')\n",
    "plt.ylabel('Importance value')\n",
    "plt.title('Feature importances comparison')\n",
    "plt.xticks(index, feature_labels)\n",
    "plt.legend()\n",
    "\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Cross Fold Validation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "'''K-Fold model evaluation'''\n",
    "folds = [3,4,5,7,10]\n",
    "kfolds=[]\n",
    "for fold in folds:\n",
    "    #with open(os.getcwd()+\"/Model/rf_pp_a_\"+str(fold)+\"fold_best.pkl\", \"rb\") as f:\n",
    "    with open(os.getcwd()+\"/Model/cfold/rf_pp_ha_a_\"+str(fold)+\"fold_best.pkl\", \"rb\") as f:\n",
    "        kfolds.append(pickle.load(f))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x504 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "ks = np.array([kfold[\"k\"] for kfold in kfolds])\n",
    "pcc_tests = np.array([kfold[\"pcc_test\"] for kfold in kfolds])\n",
    "pcc_trains = np.array([kfold[\"pcc_train\"] for kfold in kfolds])\n",
    "\n",
    "fig_size = np.zeros(2)\n",
    "fig_size[0] = 10\n",
    "fig_size[1] = 7\n",
    "plt.rcParams[\"figure.figsize\"] = fig_size\n",
    "plt.plot(ks, pcc_tests, '^-', markerfacecolor=\"None\", markeredgecolor='red', markeredgewidth=5)\n",
    "plt.title(\"K-Fold models evaluation\")\n",
    "plt.xlabel('K')\n",
    "plt.ylabel('PCC')\n",
    "#plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x504 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "total_train_data = np.array([kfold[\"train_idx\"].shape[0] for kfold in kfolds])\n",
    "total_test_data = np.array([kfold[\"test_idx\"].shape[0] for kfold in kfolds])\n",
    "plt.plot(ks, total_test_data,'o-', markerfacecolor=\"None\", markeredgecolor='red', markeredgewidth=5, label=\"Test Data\")\n",
    "plt.plot(ks, total_train_data,'o-', markerfacecolor=\"None\", markeredgecolor='blue', markeredgewidth=5, label=\"Train Data\")\n",
    "plt.xlabel(\"K\")\n",
    "plt.ylabel(\"Data\")\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Inference on various complexes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "'''Full-data model load'''\n",
    "def data_load(filename):\n",
    "    '''\n",
    "    data checker\n",
    "    '''\n",
    "    data = []\n",
    "    try:\n",
    "        with open(filename, 'rb') as fr:\n",
    "            try:\n",
    "                while True:\n",
    "                    data.append(pickle.load(fr))\n",
    "            except EOFError:\n",
    "                pass            \n",
    "    except FileNotFoundError:\n",
    "        print('File is not found')\n",
    "    return data\n",
    "\n",
    "full_model = os.getcwd()+\"/Model/rf_pp_ha_alpha_full.pkl\"\n",
    "with open(full_model, \"rb\") as f:\n",
    "    rf = pickle.load(f)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Inference on PC-PC"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=12)]: Using backend ThreadingBackend with 12 concurrent workers.\n",
      "[Parallel(n_jobs=12)]: Done  26 tasks      | elapsed:    0.0s\n",
      "[Parallel(n_jobs=12)]: Done 176 tasks      | elapsed:    0.0s\n",
      "[Parallel(n_jobs=12)]: Done 426 tasks      | elapsed:    0.0s\n",
      "[Parallel(n_jobs=12)]: Done 776 tasks      | elapsed:    0.0s\n",
      "[Parallel(n_jobs=12)]: Done 1000 out of 1000 | elapsed:    0.0s finished\n"
     ]
    }
   ],
   "source": [
    "path = os.getcwd()+\"/Data/data_PC-PC.pkl\"\n",
    "data =data_load(path)\n",
    "ids = [d[\"id\"] for d in data]\n",
    "vectors = [d[\"x_vector\"] for d in data]\n",
    "temp = ids[1]\n",
    "del ids[1]\n",
    "ids.append(temp)\n",
    "temp = vectors[1]\n",
    "del vectors[1]\n",
    "vectors.append(temp)\n",
    "pred_PCPC = rf.predict(vectors)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Inference on PC-CytF"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=12)]: Using backend ThreadingBackend with 12 concurrent workers.\n",
      "[Parallel(n_jobs=12)]: Done  26 tasks      | elapsed:    0.0s\n",
      "[Parallel(n_jobs=12)]: Done 176 tasks      | elapsed:    0.0s\n",
      "[Parallel(n_jobs=12)]: Done 426 tasks      | elapsed:    0.0s\n",
      "[Parallel(n_jobs=12)]: Done 776 tasks      | elapsed:    0.0s\n",
      "[Parallel(n_jobs=12)]: Done 1000 out of 1000 | elapsed:    0.0s finished\n"
     ]
    }
   ],
   "source": [
    "path = os.getcwd()+\"/Data/data_PC-CF.pkl\"\n",
    "data =data_load(path)\n",
    "ids = [d[\"id\"] for d in data]\n",
    "vectors = [d[\"x_vector\"] for d in data]\n",
    "temp = ids[1]\n",
    "del ids[1]\n",
    "ids.append(temp)\n",
    "temp = vectors[1]\n",
    "del vectors[1]\n",
    "vectors.append(temp)\n",
    "pred_PCCF = rf.predict(vectors)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Inference on PC-CytF experimental structure"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[6.82370286 7.34691624 7.50137857]\n",
      "['1tkw.pdb', '2jxm.pdb', '2pcf.pdb']\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=12)]: Using backend ThreadingBackend with 12 concurrent workers.\n",
      "[Parallel(n_jobs=12)]: Done  26 tasks      | elapsed:    0.0s\n",
      "[Parallel(n_jobs=12)]: Done 176 tasks      | elapsed:    0.0s\n",
      "[Parallel(n_jobs=12)]: Done 426 tasks      | elapsed:    0.0s\n",
      "[Parallel(n_jobs=12)]: Done 776 tasks      | elapsed:    0.0s\n",
      "[Parallel(n_jobs=12)]: Done 1000 out of 1000 | elapsed:    0.0s finished\n"
     ]
    }
   ],
   "source": [
    "path = os.getcwd()+\"/Data/data_PC-CF_exp.pkl\"\n",
    "data =data_load(path)\n",
    "ids = [d[\"id\"] for d in data]\n",
    "vectors = [d[\"x_vector\"] for d in data]\n",
    "pred_PCCF_exp = rf.predict(vectors)\n",
    "print(pred_PCCF_exp)\n",
    "print(ids)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Inference on 4azu"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=12)]: Using backend ThreadingBackend with 12 concurrent workers.\n",
      "[Parallel(n_jobs=12)]: Done  26 tasks      | elapsed:    0.0s\n",
      "[Parallel(n_jobs=12)]: Done 176 tasks      | elapsed:    0.0s\n",
      "[Parallel(n_jobs=12)]: Done 426 tasks      | elapsed:    0.0s\n",
      "[Parallel(n_jobs=12)]: Done 776 tasks      | elapsed:    0.0s\n",
      "[Parallel(n_jobs=12)]: Done 1000 out of 1000 | elapsed:    0.0s finished\n"
     ]
    }
   ],
   "source": [
    "path = os.getcwd()+\"/Data/data_4azu.pkl\"\n",
    "data =data_load(path)\n",
    "vectors = [d[\"x_vector\"] for d in data]\n",
    "pred_4azu = rf.predict(vectors)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "length = 10\n",
    "idx = np.linspace(1, length, length)\n",
    "fig_size = np.zeros(2)\n",
    "fig_size[0] = 15\n",
    "fig_size[1] = 5\n",
    "plt.rcParams[\"figure.figsize\"] = fig_size\n",
    "plt.xlim([0,15])\n",
    "plt.plot(idx, pred_PCPC,'o-', markerfacecolor=\"None\", markeredgecolor='red', markeredgewidth=5, label=\"PC-PC\" )\n",
    "plt.plot(idx, pred_PCCF,'o-', markerfacecolor=\"None\", markeredgecolor='green', markeredgewidth=5, label=\"PC-CytF\")\n",
    "plt.plot(11, pred_4azu,'o', markerfacecolor=\"None\", markeredgecolor='blue', markeredgewidth=5, label=\"4azu\")\n",
    "plt.plot(np.linspace(1, 3, 3), pred_PCCF_exp,'o-', markerfacecolor=\"None\", markeredgecolor='yellow', markeredgewidth=5, label=\"PC-CytF_exp; 1=1tkw, 2=2jxm, 3=2pcf\")\n",
    "plt.xlabel(\"Complex Number\")\n",
    "plt.ylabel(\"Binding Affinity\")\n",
    "plt.legend()\n",
    "plt.grid(True)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### RMSD values of docked structures against experimental structures"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "rmsds = np.loadtxt(os.getcwd()+\"/Data/rmsd\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "length = 10\n",
    "idx = np.linspace(1, length, length)\n",
    "fig_size = np.zeros(2)\n",
    "fig_size[0] = 15\n",
    "fig_size[1] = 5\n",
    "plt.rcParams[\"figure.figsize\"] = fig_size\n",
    "plt.xlim([0,11])\n",
    "plt.plot(idx, rmsds[0],'o-', markerfacecolor=\"None\", markeredgecolor='red', markeredgewidth=5, label=\"Against 1tkw\" )\n",
    "plt.plot(idx, rmsds[1],'o-', markerfacecolor=\"None\", markeredgecolor='green', markeredgewidth=5, label=\"Against 2jxm\")\n",
    "plt.plot(idx, rmsds[2],'o-', markerfacecolor=\"None\", markeredgecolor='blue', markeredgewidth=5, label=\"Against 2pcf\")\n",
    "plt.xlabel(\"Complex Number\")\n",
    "plt.ylabel(\"RMSD\")\n",
    "plt.legend()\n",
    "plt.grid(True)\n",
    "plt.show()"
   ]
  },
  {
   "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": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}