Isolation_forest.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "6c4de066",
   "metadata": {},
   "source": [
    "## Isolation Forest algorithm"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bff40f41",
   "metadata": {},
   "source": [
    "Isolation Forest, like any tree ensemble method, is built on the basis of decision trees. \n",
    "In these trees, partitions are created by first randomly selecting a feature and then selecting a random split value \n",
    "between the minimum and maximum value of the selected feature.\n",
    "In principle, outliers are less frequent than regular observations \n",
    "and are different from them in terms of values (they lie further away from the regular observations in the feature space). \n",
    "That is why by using such random partitioning they should be identified closer to the root of the tree (shorter average path length, i.e., the number of edges an observation must pass in the tree going from the root to the terminal node), with fewer splits necessary.\n",
    "\n",
    "In the following example we will use the ISO forest algorithm on a famous boston data set to detect the cities with the highest crime rate."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "d5a6704a",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.datasets import load_boston\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import warnings"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "0c5f7e8c",
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "boston =load_boston()\n",
    "names=list(boston['feature_names'])\n",
    "X=boston['data']\n",
    "df = pd.DataFrame(X, columns=names)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "7e82f2cd",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      ".. _boston_dataset:\n",
      "\n",
      "Boston house prices dataset\n",
      "---------------------------\n",
      "\n",
      "**Data Set Characteristics:**  \n",
      "\n",
      "    :Number of Instances: 506 \n",
      "\n",
      "    :Number of Attributes: 13 numeric/categorical predictive. Median Value (attribute 14) is usually the target.\n",
      "\n",
      "    :Attribute Information (in order):\n",
      "        - CRIM     per capita crime rate by town\n",
      "        - ZN       proportion of residential land zoned for lots over 25,000 sq.ft.\n",
      "        - INDUS    proportion of non-retail business acres per town\n",
      "        - CHAS     Charles River dummy variable (= 1 if tract bounds river; 0 otherwise)\n",
      "        - NOX      nitric oxides concentration (parts per 10 million)\n",
      "        - RM       average number of rooms per dwelling\n",
      "        - AGE      proportion of owner-occupied units built prior to 1940\n",
      "        - DIS      weighted distances to five Boston employment centres\n",
      "        - RAD      index of accessibility to radial highways\n",
      "        - TAX      full-value property-tax rate per $10,000\n",
      "        - PTRATIO  pupil-teacher ratio by town\n",
      "        - B        1000(Bk - 0.63)^2 where Bk is the proportion of blacks by town\n",
      "        - LSTAT    % lower status of the population\n",
      "        - MEDV     Median value of owner-occupied homes in $1000's\n",
      "\n",
      "    :Missing Attribute Values: None\n",
      "\n",
      "    :Creator: Harrison, D. and Rubinfeld, D.L.\n",
      "\n",
      "This is a copy of UCI ML housing dataset.\n",
      "https://archive.ics.uci.edu/ml/machine-learning-databases/housing/\n",
      "\n",
      "\n",
      "This dataset was taken from the StatLib library which is maintained at Carnegie Mellon University.\n",
      "\n",
      "The Boston house-price data of Harrison, D. and Rubinfeld, D.L. 'Hedonic\n",
      "prices and the demand for clean air', J. Environ. Economics & Management,\n",
      "vol.5, 81-102, 1978.   Used in Belsley, Kuh & Welsch, 'Regression diagnostics\n",
      "...', Wiley, 1980.   N.B. Various transformations are used in the table on\n",
      "pages 244-261 of the latter.\n",
      "\n",
      "The Boston house-price data has been used in many machine learning papers that address regression\n",
      "problems.   \n",
      "     \n",
      ".. topic:: References\n",
      "\n",
      "   - Belsley, Kuh & Welsch, 'Regression diagnostics: Identifying Influential Data and Sources of Collinearity', Wiley, 1980. 244-261.\n",
      "   - Quinlan,R. (1993). Combining Instance-Based and Model-Based Learning. In Proceedings on the Tenth International Conference of Machine Learning, 236-243, University of Massachusetts, Amherst. Morgan Kaufmann.\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(boston.DESCR)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "77654100",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.ensemble import IsolationForest\n",
    "\n",
    "def iso_forest_detection(df, outliers_fraction, target_feature:str, plotting = True):\n",
    "    x=df[target_feature].to_numpy().reshape(len(df[target_feature]),-1)\n",
    "    model = IsolationForest(contamination=outliers_fraction)\n",
    "    model.fit(x)\n",
    "    predictions = model.predict(x) # anomalies are -1\n",
    "    df[target_feature+'_anomalies'] = pd.Series(predictions, index = df.index)\n",
    "    decision_function = model.decision_function(x)\n",
    "    df['decision_function'] = pd.Series(decision_function, index = df.index)\n",
    "    \n",
    "    if plotting:\n",
    "        a = df.loc[df[target_feature+'_anomalies'] == -1] #anomaly\n",
    "        plt.figure(figsize=(6,6))\n",
    "        plt.plot(df[target_feature], color='blue', label= f'original {target_feature}')\n",
    "        plt.plot(a[target_feature], linestyle='none', marker='X', color='red', markersize=8, label='Anomaly')\n",
    "        plt.plot(df['decision_function'], color='green', label='decision function', alpha=0.3) # The anomaly score of the input samples.\n",
    "        #             The lower, the more abnormal. Negative scores represent outliers,\n",
    "        #             positive scores represent inliers.\n",
    "        plt.xlabel('Rows')\n",
    "        plt.ylabel(f'{target_feature}value')\n",
    "        plt.title('Anomalies')\n",
    "        plt.legend(loc='best')\n",
    "        plt.show()\n",
    "    return df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "4fc946c6",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Trying isolation forest\n",
    "\n",
    "# Import IsolationForest\n",
    "from sklearn.ensemble import IsolationForest\n",
    "# Assume that 13% of the entire data set are anomalies\n",
    " \n",
    "outliers_fraction = 0.01\n",
    "x=df['CRIM'].to_numpy().reshape(506,-1)\n",
    "model =  IsolationForest(contamination=outliers_fraction)\n",
    "model.fit(x)\n",
    "predictions = model.predict(x) # anomalies are -1\n",
    "# predictions = predictions.reshape(2375,-1)\n",
    "df['CRIM_anomalies'] = pd.Series(predictions, index = df.index)\n",
    "decision_function = model.decision_function(x)\n",
    "df['decision_function'] = pd.Series(decision_function, index = df.index)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "8541aaf0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df['CRIM'].hist(bins=40);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "9e355b74",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAX4AAAEWCAYAAABhffzLAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuNCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8QVMy6AAAACXBIWXMAAAsTAAALEwEAmpwYAABDmElEQVR4nO2deXhV1dX/PysDhDCDiChKUCxTSEJAGZSpWnCgIgoOKAKKaEd9nYdWsa2v/tS3amurtSpQtWCLBa1Dq2gQnFDAWUBRgqLMMySB5N71+2PfkzvkJrkZbkLuXZ/nuc85+wz7rH2H71l37X3WFlXFMAzDSB5SGtsAwzAMo2Ex4TcMw0gyTPgNwzCSDBN+wzCMJMOE3zAMI8kw4TcMw0gyTPgNo4EQkZki8lRg/RgR2SciqY1tl5F8mPAbCYWILBaRnSLSvLFtqQpV/UZVW6mqr7FtMZIPE34jYRCRLGAYoMBZjWuNYRy6mPAbicQlwLvAbGCKt1FEZovIn0TkRRHZKyLLROS4kP1DReR9EdkdWA4N2bdYRH4nIm8HQjP/FpGOIvK0iOwJHJ8VcvyDIvJtYN8KERkWzVARyRIRFZG0QLmtiDwuIhtF5LvANVMD+3qIyBsB+7aJyDP1/L4ZSYYJv5FIXAI8HXiNEZHOIfsuBO4A2gNrgTsBRKQD8CLwB6Aj8HvgRRHpGHLuBcBk4CjgOOAdYBbQAVgF3B5y7PtAXmDf34F/ikhGDLbPAcqAHkB/YDQwPbDvt8ArAdu7An+MoT7DqBQTfiMhEJGTgW7AP1R1BfAVMCnkkH+p6nuqWoa7MeQFtp8JfKmqT6pqmarOBVYDPw45d5aqfqWqu4GXga9UdVGgrn/ihBoAVX1KVbcH6vo/oDnQsxrbOwOnA1er6n5V3QLcj7vhAJQG2nakqpao6ps1fX8MIxQTfiNRmAK8oqrbAuW/ExLuATaFrBcBrQLrRwLrI+paj/PuPTaHrBdHKXt1ISLXisiqQFhmF9AWOKwa27sB6cBGEdkVOO8vwOGB/TcAArwnIp+JyKXV1GcYVZLW2AYYRl0RkRbAeUCqiHgC3xxoJyK51Zz+PU54QzkG+E8t7BgG3AicAnymqn4R2YkT7ar4FjgAHBb4FxGGqm4CLg9c42RgkYgsUdW1NbXRMMA8fiMxOBvwAX1wIZw8oDewFBf3r4qXgB+IyCQRSROR8wP1vFALO1rj4vRbgTQRuQ1oU91JqroRF8P/PxFpIyIpInKciIwAEJGJItI1cPhO3KglGwZq1BoTfiMRmIKLw3+jqpu8F/AQcBFV/LNV1e3AWOBaYDsurDI2JGRUE/6L6wP4AhcuKsF587FwCdAM+Bwn7vOBLoF9JwDLRGQf8Dxwlaquq4V9hgGA2EQshmEYyYV5/IZhGEmGCb9hGEaSYcJvGIaRZJjwG4ZhJBlNYhz/YYcdpllZWY1thmEYRpNixYoV21S1U+T2JiH8WVlZLF++vLHNMAzDaFKISORT6YCFegzDMJIOE37DMIwkw4TfMAwjyWgSMf5olJaWsmHDBkpKShrbFMMwGoiMjAy6du1Kenp6Y5vSpGmywr9hwwZat25NVlYWItUlPzQMo1HZswcKCyErC9q0qViOAVVl+/btbNiwge7du8fR2MSnyYZ6SkpK6Nixo4m+YRzq7NkDa9fCwYNuuWlTeHnPnpiqERE6duxo//LrgSYr/ICJvmE0BQoLwe93634/fP99eLmwMOaq7DdfPzRp4TcMowmQlQUpIVLjiT647Ra2aXBM+BuAM844g127dlV5zG233caiRYtqVf/ixYsZO3Zs1H3vvfcew4cPp2fPnvTq1Yvp06dTVFTE7Nmz6dSpE3l5efTq1Yv777+//JyZM2dy3333ATB16lQyMzPZu3dv+f6rrroKEWHbttqkrDeSjjZt4Mgjw8UfXPnII6F168axK4lpsp27TQFVRVV56aWXqj32N7/5Tb1ff/PmzUycOJF58+YxZMgQVJVnn322XMTPP/98HnroIbZv307Pnj2ZMGECRx99dIV6evTowXPPPcfFF1+M3++noKCAo446qsJxhhGVPXvCwzseXtinZUsT/wbGPP468Pvf/57s7Gyys7N54IEHACgsLKR379789Kc/JT8/n2+//ZasrKxy7/i3v/0tvXr14kc/+hEXXnhhmGc9f/58wKWouP3228nPz6dfv36sXr0acN770KFD6d+/P0OHDmXNmjVV2venP/2JKVOmMGTIEMDFRydMmEDnzp3DjuvYsSM9evRg48aNUeu58MILeeaZZwD37+Kkk04iLc18BiNGQmP8gI+IsM86m0ysoUmIX+/VV8OHH9ZvnXl5ENDyqKxYsYJZs2axbNkyVJVBgwYxYsQI2rdvz5o1a5g1axZ//vOfw85Zvnw5zz77LB988AFlZWXk5+czYMCAqPUfdthhrFy5kj//+c/cd999PPbYY/Tq1YslS5aQlpbGokWLuOWWW3j22WcrtfHTTz9lypQp1bb1m2++oaSkhJycnKj7jz/+eJ577jl27tzJ3Llzufjii3n55ZerrdcwABfjX7sW/H58pPA9R3J0SuAfQEqK2280KObx15I333yT8ePH07JlS1q1asU555zD0qVLAejWrRuDBw+Oes64ceNo0aIFrVu35sc//nGl9Z9zzjkADBgwgMLAqIfdu3czceJEsrOz+Z//+R8+++yzOrXhmWeeoW/fvhx77LFcddVVZGRkVGnPvHnzWLZsGcOGDavTdY0ko00b6NEDmjVjLcezmSPKy/ToEfM4fqP+SAiPvyrPPF5UNVdxy5Yta3xOJM2bNwcgNTWVsrIyAH79618zatQoFixYQGFhISNHjqyyjr59+7JixQrGjRsXdb8X43/nnXc488wzOf300zniiCOiHnvBBReQn5/PlClTSInspDOM6mjTBnJy2Ls8vGw0DvYLriXDhw9n4cKFFBUVsX//fhYsWFCtJ3zyySfz73//m5KSEvbt28eLL75Yo2vu3r27vFN19uzZ1R7/85//nDlz5rBs2bLybU899RSbNm0KO27IkCFMnjyZBx98sNK6jjnmGO68805++tOf1shmwzAOPUz4a0l+fj5Tp07lxBNPZNCgQUyfPp3+/ftXec4JJ5zAWWedRW5uLueccw4DBw6kbdu2MV/zhhtu4Oabb+akk07C5/NVe3znzp2ZN28e1113HT179qR3794sXbqUNlH+Wt94443MmjUrbNhmJFdccQXHHXdczPYahnFoIjUJPzQWAwcO1MiJWFatWkXv3r0byaLas2/fPlq1akVRURHDhw/n0UcfJT8/v7HNMowGwfsZDxxY+zqa6m+/MRCRFapa4d1OiBh/U2LGjBl8/vnnlJSUMGXKFBN9wzAaHBP+Bubvf/97Y5tgGEaSYzF+wzCMJMOE3zAMI8kw4TcMw0gyTPgNwzCSjOQQ/oIClw+koCB6uQ4sWLAAESlPpNZYtGrVKq71e6mld+3aFZaD6Pvvv2fChAlxvTaEp61+4IEHKCoqKt8XS9sXL17M22+/XV4OTYpXGyq75iOPPMLf/va3Ks+dPXs2P//5z2t9bY+FCxfy+eef16mOWFKGx5P//d//bbRrJzOJL/wFBTB2LKxf75b33BNerqP4z507l5NPPpl58+bVk8GHJi+99BLt2rWrIPxHHnlknQQ0Vn7zm99w6qmnAhWFPxYihT9eXHnllVxyySVxvw7UTfhVFb/fX/65xovKHjT0Hh8y4W8kvJzx8XgB/wN8BnwKzAUygA7Aq8CXgWX76uoZMGCARvL5559X2BaVbt1U3ffMvTIzw8vdusVWTxT27t2rRx55pK5Zs0Z79uxZvr2goEBHjBih5557rvbs2VMnTZqkfr9fVVUXLVqkeXl5mp2drdOmTdOSkpKAmd305ptv1sGDB+uAAQN0xYoVOnr0aD322GP14YcfLr/eD3/4Q+3fv79mZ2frwoULy6/ZsmVLVVW9+OKLw7ZPmjRJn3vuuRq1aerUqZqdna39+vXT+fPnl9u3detWPf/88zUjI0Nzc3P1uuuu03Xr1mnfvn1VVbWsrEyvu+46HThwoPbr108feeQRVVX9/vvvddiwYZqbm6t9+/bVJUuWhF1z2bJlOn78eFVVXbhwoWZkZOiBAwe0uLhYu3fvrqqqU6ZM0X/+85/64IMPanp6umZnZ+vIkSPL237LLbdoTk6ODho0SDdt2hRW/7p167Rz58565JFHam5uri5ZskSnTJmiv/jFL3TIkCHavXt3/ec//1l+/D333FPehttuuy3q+1TZNW+//Xa99957VVX1vffe0379+ungwYP1uuuuK3+fZs2apePHj9cxY8Zojx499Prrry+v97///a8OHjxY+/fvrxMmTNC9e/eqquqNN96ovXv31n79+um1116rb731lrZv316zsrI0NzdX165dG2bfpk2b9Oyzz9acnBzNycnRt956S9etW6e9evXSn/zkJ5qXl6eFhYXln+u6deu0Z8+eetlll2nfvn110qRJ+uqrr+rQoUO1R48eumzZMlVV3bdvn06bNk0HDhyoeXl5Yd81j4KCAh05cqReeOGF2rt3b1VVHTdunObn52v37n305pv/on6/a1NKSorm5ubqpEmTVFX1ySef1BNOOEFzc3N1xowZWlZWVqH+mH/7hgLLNZo2R9tYHy/gKGAd0CJQ/gcwFbgHuCmw7Sbg/1VXV52E//XXK4p96E2goCDmNzGSJ598Ui+99FJVVR0yZIiuWLFCVd0Xv02bNvrtt9+qz+fTwYMH69KlS7W4uFi7du2qa9asUVXVyZMn6/3336+qTlj//Oc/q6rq1Vdfrf369dM9e/boli1btFOnTqqqWlpaqrt371ZV1a1bt+pxxx1XfkPxhH/x4sU6btw4VVXdtWuXZmVlaWlpacxtuuGGG/Sqq64qL+/YsaPcPk8gPAFT1bDyX/7yF/3tb3+rqqolJSU6YMAA/frrr/W+++7T3/3ud6rqbg579uwJu2ZpaalmZWWpquq1116rAwcO1DfffFMXL16sF1xwgaoGhT/UFg9An3/+eVVVvf7668ttCCVUkL36JkyYoD6fTz/77DM97rjjVNUJ7+WXX65+v199Pp+eeeaZ+sYbb1Sor7Jrhl6nb9+++tZbb6mqE7lQ4e/evbvu2rVLi4uL9ZhjjtFvvvlGt27dqsOGDdN9+/apqurdd9+td9xxh27fvl1/8IMflH/WO3furPCeRHLeeeeVf7fKysp0165dum7dOhURfeedd8qPC/1cU1NT9eOPP1afz6f5+fk6bdo09fv9unDhwvLv1M0336xPPvlkuR3HH398ub0eBQUFmpmZqV9//XX5tu3bt6uq6tKlRXrssX1169Ztqhr83qq63/TYsWP14MGDqqr6k5/8ROfMmVOhbSb8sVOZ8Mc71JMGtBCRNCAT+B4YB8wJ7J8DnB1XC0aNgttvh8zM8O2ZmTBzJlST4bIq5s6dywUXXAC47JVz584t33fiiSfStWtXUlJSyMvLo7CwkDVr1tC9e3d+8IMfADBlyhSWLFlSfs5ZZ50FQL9+/Rg0aBCtW7emU6dOZGRksGvXLlSVW265hZycHE499VS+++47Nm/eHGbTiBEjWLt2LVu2bGHu3Lmce+65NZo0ZdGiRfzsZz8rL7dv3z7mc1955RX+9re/kZeXx6BBg9i+fTtffvklJ5xwArNmzWLmzJl88skntI6YbSktLY0ePXqwatUq3nvvPa655hqWLFnC0qVLY0oB3axZs/KpJ0PTWFfH2WefTUpKCn369Cl/H1955RVeeeUV+vfvT35+PqtXr+bLL7+s8TV37drF3r17GTp0KACTJk0K23/KKafQtm1bMjIy6NOnD+vXr+fdd9/l888/56STTiIvL485c+awfv162rRpQ0ZGBtOnT+df//oXmZHf5Si8/vrr/OQnPwFchlcvJ1RlKcMBunfvTr9+/UhJSaFv376ccsopiAj9+vUrb98rr7zC3XffTV5eHiNHjqSkpIRvvvmmQl0nnngi3UPm0v3DH/5Abm4ul146mM2bv436nr722musWLGCE044gby8PF577TW+/vrrattq1Jy4Pbmrqt+JyH3AN0Ax8IqqviIinVV1Y+CYjSJyeLTzRWQGMANcZshaU1AAd9wBkTHhoiIn/CecUCvx3759O6+//jqffvopIoLP50NEuOeee4BgWmUIplbWavIieeekpKSEnZ+SkkJZWRlPP/00W7duZcWKFaSnp5OVlUVJSUmFeiZPnszTTz/NvHnzeOKJJyrsv/XWW8szg34YMYONqiIisb0JEagqf/zjHxkzZkyFfUuWLOHFF19k8uTJXH/99RXi4MOGDePll18mPT2dU089lalTp+Lz+cpnKKuK9PT0cptD01hXR+h77H02qsrNN9/MFVdcUadrxvpZh56vqvzoRz8KcyA83nvvPV577TXmzZvHQw89xOuvv1514yqhspThkTaFfge97x9QPn1nz549Y77O4sWLWbRoEe+88w6ff57JFVeMpLi44vdWVZkyZQp33XVXjdpk1Jy4efwi0h7n3XcHjgRaisjFsZ6vqo+q6kBVHdipU6faGzJtWrjoh3pLRUUwdWqtqp0/fz6XXHIJ69evp7CwkG+//Zbu3bvz5ptvVnpOr169KCwsZO3atQA8+eSTjBgxIuZr7t69m8MPP5z09HQKCgpYv3591OOmTp1aPhVk3759K+y/8847+fDDDyuIPsDo0aN56KGHyss7d+4M29+6detKM3iOGTOGhx9+mNLSUgC++OIL9u/fz/r16zn88MO5/PLLueyyy1i5cmWFc4cPH84DDzzAkCFD6NSpE9u3b2f16tVR7a/KhsqI9ZwxY8bwxBNPsG/fPgC+++47tmzZUqNrgfun1Lp1a959912AmDr/Bw8ezFtvvVX+/SgqKuKLL75g37597N69mzPOOIMHHnig/HOrqk2nnHIKDz/8MOA6WPfs2VPjNkRjzJgx/PGPfyy/sX3wwQfVnrN7927at29PZmYmhYWr+fTTd8v3paenl39fTjnlFObPn1/+fu/YsaPS77hRN+IZ6jkVWKeqW1W1FPgXMBTYLCJdAALLmv+qasKsWUGx98I7oeVZs2pV7dy5cxk/fnzYtnPPPbfKXDwZGRnMmjWLiRMnlv+lvvLKK2O+5kUXXcTy5csZOHAgTz/9NL169Yp6XOfOnenduzfTpk2LuW6PX/3qV+zcuZPs7Gxyc3MpiBj11LFjR0466SSys7O5/vrrw/ZNnz6dPn36kJ+fT3Z2NldccQVlZWUsXryYvLw8+vfvz7PPPstVV11V4bqDBg1i8+bNDB8+HICcnBxycnKi/vuYMWMGp59+OqNGjYq5XT/+8Y9ZsGABeXl55TOlRWP06NFMmjSJIUOG0K9fPyZMmFDjm4zH448/zowZM8onuq8uBXenTp2YPXs2F154ITk5OQwePJjVq1ezd+9exo4dS05ODiNGjOD+++8HXHjx3nvvpX///nz11VdhdT344IMUFBTQr18/BgwYUOfZ2jx+/etfU1paSk5ODtnZ2fz617+u9pzTTjuNsrIycnJyeOSRX5OdHQw1zZgxg5ycHC666CL69OnD7373O0aPHk1OTg4/+tGPKp0H2qgbcUvLLCKDgCeAE3ChntnAcuAYYLuq3i0iNwEdVPWGquqqc1rmggLn+c+e7cI6XnnWLNcHkGAUFRXRr18/Vq5cWaN8/0b94qXgBrj77rvZuHFjlZPdJAPezzg/3023WxssLXPsNHhaZlVdJiLzgZVAGfAB8CjQCviHiFyGi/9PjJcN5YwaBaGdb5HlBGLRokVceumlXHPNNSb6jcyLL77IXXfdRVlZGd26dYtp1jTDaAhsIhbDMBoM72fcvz+kptauDvvtx05lHn/iP7lrGIZhhGHCbxiGkWSY8BuGYSQZJvyGYRhJhgl/PTFz5syYnjKNhvdYf2XUV+rcrVu3MmjQIPr371/lWPaaMnv2bL7//vvy8vTp0+ucLthIbJrAmJKExoT/EKC6dMH1lTr3tddeo1evXnzwwQcx5cCJlUjhf+yxx+jTp0+91W8YVRLH+TYSFRP+OnDnnXfSs2dPTj31VNasWVO+/auvvuK0005jwIABDBs2rHySls2bNzN+/Hhyc3PJzc0tF3zvIZ+NGzcyfPhw8vLyyM7OLvfKs7Ky2LZtGwC///3vyc7OJjs7uzwtQ2FhIb179+byyy+nb9++jB49muLi4jBbP/zwQ2644QZeeukl8vLyKC4uDptMZP78+UwNpK+YOnUqv/zlLxk6dCjHHntsWL79e+65h379+pGbm8tNN93E/PnzWb58ORdddFF5vSNHjsQbfjt37lz69etHdnY2N954Y3k9rVq14tZbbyU3N5fBgwdXSDZnGDER5/k2EpW4PcDVkHy25TN2H9hdr3W2bd6WvodXzBPjsWLFCubNm8cHH3xAWVkZ+fn5DBgwAHCPoT/yyCMcf/zxLFu2jJ/+9Ke8/vrr/PKXv2TEiBEsWLAAn89Xng/G4+9//ztjxozh1ltvxefzVZhsZMWKFcyaNYtly5ahqgwaNIgRI0bQvn17vvzyS+bOnctf//pXzjvvPJ599lkuvjiYGikvL4/f/OY3LF++PCwXT2Vs3LiRN998k9WrV3PWWWcxYcIEXn75ZRYuXMiyZcvIzMxkx44ddOjQgYceeoj77ruPgQPDhwt///333HjjjaxYsYL27dszevRoFi5cyNlnn83+/fsZPHgwd955JzfccAN//etf+dWvflWtXYYRRmgurqIidOYdSHGwzLRpCfuwZl1ICOFvDJYuXcr48ePLU+R6KZX37dvH22+/zcSJwQeSDxw4ALhUud60fKGpcj1OOOEELr30UkpLSzn77LPJy8sL2//mm28yfvz48syH55xzDkuXLuWss86ie/fu5cfXJDVxZURLWbxo0SKmTZtW3uYOHTpUWcf777/PyJEj8ZLsXXTRRSxZsoSzzz67QlrjV199tU72GknKrFnOsw+If7nog8vFZU9LRyUhhL8qzzyeREsg5vf7adeuXdTMl9UxfPjwKtMXV/WUdWSa38hQTzRC7Y9M71xZyuKapGyuyt7aplI2jDC8+TYiU6/Xw3wbiYzF+GvJ8OHDWbBgAcXFxezdu5d///vfALRp04bu3bvzz3/+E3Di99FHHwHVp8qtLn3x8OHDWbhwIUVFRezfv58FCxbUqZO2c+fOrFq1Cr/fz4IFC6o9fvTo0TzxxBPlIagdO3YAlacHHjRoEG+88Qbbtm3D5/Mxd+7cGqWhNhKXehvVU918G4sX19OFEgsT/lqSn5/P+eefT15eHueee26YAD/99NM8/vjj5Obm0rdvX5577jmg+lS51aUvzs/PZ+rUqZx44okMGjSI6dOn079//1q34e6772bs2LH88Ic/pEuXLtUef9ppp3HWWWcxcOBA8vLyyoevTp06lSuvvLK8c9ejS5cu3HXXXYwaNYrc3Fzy8/MZN25cre01jApEzLexn/qZbyPRsSRthmE0GN7PODcX0tNrV0fYb98b1VNUxH4ymclM7s2c6UQ/MxNeeCEhU6/HiiVpMwwj8Rg1yol7t26cyYvcx/Xl5WQX/apIiM5dwzCSmMD8Gm9IeNmonCbt8TeFMJVhGBWp7U/XfvP1Q5MV/oyMDLZv325fBMNIElSV7du3k5GR0dimNHmabKina9eubNiwga1btza2KYZhxEgg8whr1kBarOpTUuJOPOwwMtq1o+v69S6ck6BzZjcETVb409PT6d69e2ObYRhGDfBy9337LXTtGsMJIaN2yMwMf1hr7FjrwK0lTTbUYxhG0yXmCG1ELp6wh7W8XDxGjTHhNwyjwYlZ+GfNcp6+R2RaBsvFUytM+A3DaHBiFn4vF0+o+IPl4qkjJvyGYTQ4MQu/5eKJCyb8hmE0OLWK8WO5eOoLE37DMBqc2sT4vVw86oV9MjPdfqPGmPAbhtHg1CjGH5GL5+CzlounrjTZcfyGYTRdavTAfUQunrJho2huuXjqhHn8hmE0OHXJtGJZWuqOCb9hGE0Kv7+xLWj6mPAbhtHgmMffuJjwG4bR4NRFvM3jrzsm/IZhNDjm8TcuJvyGYTQ45vE3Lib8hmE0OObxNy4m/IZhNDjm8TcucRV+EWknIvNFZLWIrBKRISLSQUReFZEvA8v28bTBMIxDD/P4G5d4e/wPAv9R1V5ALrAKuAl4TVWPB14LlA3DSCLM429c4ib8ItIGGA48DqCqB1V1FzAOmBM4bA5wdrxsMAzj0MQ8/sYlnh7/scBWYJaIfCAij4lIS6Czqm4ECCwPj6MNhmEcgpjH37jEU/jTgHzgYVXtD+ynBmEdEZkhIstFZPnWrVvjZaNhGI2AefyNSzyFfwOwQVWXBcrzcTeCzSLSBSCw3BLtZFV9VFUHqurATp06xdFMwzAaGhP+xiVuwq+qm4BvRaRnYNMpwOfA88CUwLYpwHPxssEwjEMTC/U0LvHOx/8L4GkRaQZ8DUzD3Wz+ISKXAd8AE+Nsg2EYhxjm8TcucRV+Vf0QGBhl1ynxvK5hGIc25vE3LvbkrmEYDU48PX77R1A9JvyGYTQ48fT47R9B9ZjwG4bR4JjH37jEJPwi0iJkdI5hGEadiKfHb8JfPdUKv4j8GPgQ+E+gnCciz8fZLsMwjKiYx193YvH4ZwInArugfKROVrwMMgwj8TGPv3GJRfjLVHV33C0xDCNpsBh/4xLLOP5PRWQSkCoixwO/BN6Or1mGYSQy5vE3LrF4/L8A+gIHgLnAHuDqONpkGEaCYx5/41Ktx6+qRcCtgZdhGEadsXH8jUsso3oKROT1yFdDGGcYRmJSW+EfSQHZY7OgoMBtKCiArJByHepOJmKJ8V8Xsp4BnAuUxcccwzCSgdqI80gKeIGxNN9UBGPHwu23wx13QJErj+QFFjPKhD8GYgn1rIjY9JaIvBEnewzDSAJqI86zmEZLilyhqCgo+oHyLKbRnUIT/hiIJdTTIeR1mIiMAY5oANsMw0hQaiPO05jFfjKDGzzRB8jMZCqza113shHLqJ4VwPLA8h3gWuCyeBplGEZiUxtxXswo7uB2fBmZ4TsyM2HmTN5gZK3rTjZiCfV0bwhDDMNIHmob47+dO0gtKQrfUVQEM2cyghN4g5Em/DFQqfCLyDlVnaiq/6p/cwzDSAbqHOMH9pMZFvOfzVSL8cdIVR7/j6vYp4AJv2EYtaK2Mf4XGEtLithPJjOZyT0tZiLFRZCZybSiWYCN44+FSoVfVac1pCGGYSQPtY3xj+UFXu48jXN2zuaVgyOZ+Y+BtPz5NJg1i8U/HFXrupONmObcFZEzcWkbMrxtqvqbeBllGEZiU1txXswo3plbyNIzXbn05FFQWFgvdScTsQznfAQ4H5ezR4CJQLc422UYRgJT15QNIpXXY8JfPbEM5xyqqpcAO1X1DmAIcHR8zTIMI5GpryRt0eL5JvzVE4vwFweWRSJyJFAK2BBPwzBqTX0kaRtJAe3yssrz9IykgHVkkf5mQeUnG0Bswv+CiLQD7gVWAoW49MyGYRi1oq4e/3Cfy9uTumG9y9tzzz28wFiyWE+HKWPDkrYZFYnlAa7fBlafFZEXgAybkcswjMbC74dHDlbM2+OVU4qLYNq0Cp2+RpBYOnc/EpFbROQ4VT1gom8YRl2pq8d/ZbPK8/b4W2TC7Nm1v0ASEEuo5yxcGuZ/iMj7InKdiBwTZ7sMw0hgair8kR26S1Kj5+3ZTya7r54JI0fW2cZEplrhV9X1qnqPqg4AJgE5wLq4W2YYRsJSU+EPHb2jCsPKouftaUkRbR+YCYsX19nGRCYWjx8RyRKRG4B5QC/ghrhaZRhGQlMX4ff74S8Hw/P2FEnQ808pLoKpU+toYWITS4x/GS4vTwowUVVPVNX/i7tlhmEkLDUVfp8v/Nwrmwdj/PvJ5IG2M8vL/oxMmDWrvkxNSGLx+Keoar6q3q2qX8fdIsMwEp66Cv+SVJe3p6hTN87kRR5tez1jeYFCurHxry/AqFH1a3CCEUuMf3VDGGIYRvJQ11APuLw9r/61sHwClsWMojuFFA820a+OmGL8hmEY9UmNhL+ggJZ9sxiJeyjrsE8K+LzYlcvK6lh3kmLCbxhGgxOzOBcUwNixpHy7nhcYy/Xcw0l3jaWbunL7Dys+oWv5+KvHZuAyDKPBiVn4p00rfzirJUXczh2kHQyWT3x4Gi6LTC3qTmKqStkwH/gw8AKXktnDZuAyDKPWxCzOs2a5XDwh4u+xn0wWT50NEWMMTfirp6pQz7nAFwQf2LpTVacFXpfGegERSRWRDwJ5fhCRDiLyqoh8GVi2r1MLDMNocsQszqNGwe23uzQMIXhTL67vPhII5uevUd1JTKXCr6oLVPUCYATwFfB/IvKmiIyo4TWuAlaFlG8CXlPV44HXAmXDMJKIGsX477jDPZQVQkuKmMlMOq9aXPu6k5hYOndLgN3AHqAlIdMvVoeIdAXOBB4L2TwOmBNYnwOcHWt9hmEkBrWJ8QNhidlaUsSPnp5a+7qTmEqFX0RGicijwApgFPCgqvZX1f/WoP4HcOkdQvvZO6vqRoDA8vBKrj9DRJaLyPKtW7fW4JKGYRzq1CjGnxl8QncmMylrFiw/P77iE7om/NVTlcf/GnAi8CbQHLhERP7gvaqrWETGAltUdUVtDFPVR1V1oKoO7NSpU22qMAzjEKVGMf4XXuBgF/eE7n1cz6KrX2C9dGMsL/Bl14oPa5nwV09Vo3ouxY3eqS0nAWeJyBm48FAbEXkK2CwiXVR1o4h0AbbU4RqGYTRBYhHn0lI3Jr/5qFF8+kIhbwxw27/vOYpzWxRSVASDDlY8z8bxV0+lwq+qs+tSsareDNwMICIjgetU9WIRuReYAtwdWD5Xl+sYhtH0iEX4jz0WNmxwx4Y+oRt67sEowm8ef/VU9QDXv6nC41fVs2p5zbtxk7pcBnwDTKxlPYZhJDAbNgTXQ4U/1KOvTvgPHoS77oIbboAWLerfxqZKVaGe++rrIqq6GFgcWN8OnFJfdRuG0fSoqVdeWhp+rnd+dcL/yCMwc6Zbv/32ml0zkakq1PNGZftE5KT4mGMYRjJQU+GvrcdfVBS+NBxVhXpSgfOAo4D/qOqngZE6twAtgP4NY6JhGIlAqCDXRfhrE+MPfbLXqDrU8zhwNPAe8AcRWQ8MAW5S1YUNYJthGAlEXYQ/MtTjCbkn/JWlbLCO3uhUJfwDgRxV9YtIBrAN6KGqmxrGNMMwEon68vhrEurxMI8/nKoe4Dqoqn4AVS0BvjDRNwyjttSnx+8RTfhDbwzm8UenKo+/l4h8HFgX4LhAWQBV1Zy4W2cYRsJgHv+hQ1XC37vBrDAMI+FpzM5dI5yqhD8dl1DtrdCNIjIM+D6uVhmGkXDUV6inJh6/3QSiU1WM/wFgb5TtxYF9hmEYMVOfHn+sD3B5WKgnnKqEP0tVP47cqKrLgay4WWQYRkISjxi/90/A56t93clIVcJf1YQrlvXCMIwaEc9RPZX1AdhNIDpVCf/7InJ55MZAcrVa5dg3DCN5ieeonuqE30I94VTVuXs1sEBELiIo9AOBZsD4ONtlGEaCUZ8xfk/IDxyouD9aPn4T/nCqStK2GRgqIqOA7MDmF1X19QaxzDCMhCKeo3oq8/iN6FTl8QOgqgVAQQPYYhhGAhPPcfwW468ZVcX4DcMw6o3G8Pgtxh8dE37DMBqEunr8zZoFz/XO924IFuqpGSb8hmE0CHUV/vR0tx6t89YSs9UME37DMBqEughyaWlQ+Kurx0I91WPCbxhGg1Bbj1816PGnpNRM+I3omPAbhtEg1Fb4/X4n/GlpznOPFuqJPN6oGhN+wzAahLoIf2mpE37P468qdGOhnuox4TcMo0GorfD7fMFQTywev2XnrB4TfsMwGoT6CPVYjL9+MOE3DKNBqI9QT009fm/dbgbhmPAbhtEg1Eeox/P4qzo/dJ93kzDhD8eE3zCMBqG2wv+nP9VsVE9o3d4ELaETtRgm/IZhNBDRhH/v3vA8PNG49VaXfrk2MX7vJmFDPMMx4TcMo0GIJvxt2sD4GGb32L8/9lE9ofs9T9+EPxwTfsMwGoR33w2uh94EXnyx+nP376+dx2+hnuiY8BuG0SBMnBhcV61ajCPFvagoPMYf6wNc5vFHx4TfMIwGRxWKi4NlbwrF0P2heKEei/HXDyb8hmE0OJHCv359+P7IfwNeqMdG9dQPJvyGYTQ4qi5847FvX/j+SHEP7dyNHMefEqFiNQn1PPecq3P37prZ39Qx4TcMo8GJ9PhLSsL3Rwq1zxfeuRsq7i1aVKw7sp7KhP93v3PLL76I3fZEIG7CLyJHi0iBiKwSkc9E5KrA9g4i8qqIfBlYto+XDYZhHJpECn/oOkQX6tBQT6i4Z2ZWrNvDQj3RiafHXwZcq6q9gcHAz0SkD3AT8JqqHg+8FigbhpFEVOfxRxNqr3PX7w+/MUR6/LUZx59sKR3iJvyqulFVVwbW9wKrgKOAccCcwGFzgLPjZYNhGIcmfj/cdluwXF2oB5zH36wZHDwYvj8Wj99G9YST1hAXEZEsoD+wDOisqhvB3RxE5PBKzpkBzAA45phjGsJMwzDiwKpV8Omn4duWLoUlS4LlWEM9GRnuJlGVx2/DOasn7sIvIq2AZ4GrVXWPxDgjgqo+CjwKMHDgwCT7I2YYiUOfPhW3NWsWXo7F409PDwp/fcX4PTmqLl9QohHXUT0iko4T/adV9V+BzZtFpEtgfxdgSzxtMAyj8agsdp6aGl6O9PijCXWoxx9KVR5/rKGegwer3p9oxHNUjwCPA6tU9fchu54HpgTWpwDPxcsGwzAal6+/jr498kndWGP8GRnh4//BJXoLpTbCn2wefzxDPScBk4FPROTDwLZbgLuBf4jIZcA3wMTopxuG0dTZsSP69kgPuyahnp07w7d36BBejhbjr244Z7J5/HETflV9E6gsoH9KvK5rGMahQ2QIxyPS469JqCfS4+/cObxsoZ7qsSd3DcOIG5Eifc890Lx5uPC3alUzjz+yziOOqPxcC/VEx4TfMIy4EenJe/l2QoW/RYuaDeeMFP72Ec/+1yTU443qMY/fMAyjnogUaU/4PaF9//3oI3VqIvyVDecsK4Nvv628vlDM4zcMw6gnonn8qanB7W3bRhf+ylI21GQ45w03wOrVbj2eMf41a+Cxx2p/fmNgwm8YRtyIJvwtWsCePa7crFnNQz2RVObxv/xy5fXt3g3Tp8Pmza5cF+EfNQouv7xih/Vbb1W8SR0qmPAbhhE3ooV6MjODgliZF18T4a/M4w/dHvkP4t574fHHgxPA1CXU4920vv8+uG39ejj5ZLjyytrXG09M+A3DiBvRPP5QD70mHr93k4ikMo8/dHtkfUuXhpfr4vEfHsg2tmFDcJt3w5szB2bPrn3d8cKE3zCMuFFUFC7W6enQsmWw3KyZE+jIfwZVjeMP5YgjYvP4PeFXhZyc8ARxUDfh79TJLUOFPzQl2bRpta87XpjwG4ZRbzz1lBO9bdtcubjYCbCXmyeax9+ypZtaEZzg33ADfPNNxbojhf/ee13nbaTH74l8tFDPgQPwyScV665LqCeax3+oDw814TcMo9744x/dcu1atywudsLcvLkrR3r8XtkT/nffdYIezUuODPX06OFGBVUW6klPD27zbgahIaVQr/zgQXj99WBnb03w2vbdd8FtkR29hxom/IZh1BuemHriW1TkPG9PsEM9/pQU90+gZcvgZOueZx45+TpU9Pi9eioL9YR2GHvCH7rt+OOD60VFcMopMHJktU2sgOfdezcvMOE3DCOJiBR+z+OPJvxeTv5WrYKi6YloLDF+r55Qzx5cErcLLwz3wL36Qj3+H/wguO6Fabxx/zXBszn0pnKoC3+DzMBlGEZy4Am/52F7Hn+0UI8n/C1bOqH87jsn2KHnh5KeHi7ykSEe7/q//W2wnJPjrl+dxx+tTyFWPJFvSsJvHr9hGPVGZO4br3M3msfvibh3I7jmmmCncLQJXNLSXEzfI5rwnxKR9/e446Bjx+jCf9RRwfW6CH9T9PhN+A3DqDc84fdEcOfOYFoGj0jhb9XKLasbCVNWFi78kbF9gDPPDC9nZrq+hMhQT//+MHRo8Ljt24Pru3dXbUck0YQ/tC3R7GxsTPgNw6g3IoV/0ybo0iUY6jlwIOjhe169V65uOu79+yv3+D/4wD0te9hh4edkZroOZL8fvvgiONrowQcrTuDisXFj1XZEUp3HfygKv8X4DcOoN0KF/+BBF7rp0iUYSvH5goLtjZ2PvBFURlFR5cKfl+eWkSmaW7RwHr/fDz17BrdnZEC7dtGv07s3PP00TJoUvG7z5hXnCfaoLsafcgi614egSYZhNFVChf9Pf3LrRxwBTzwBt9zi8tdECr8X6omcUhHgpptg1iy3np0dHjKK5klHCn9mprMp8qGtFi0qF36Aiy5yS5/P3Zh+/vPKj63K4588+dBM1GbCbxhGjfC89kcfrbjPE/49e1xnLTjh79wZ7rzTeb8dO7rte/e6pefxb9lSsb5x42DqVNi1C/Lzw/dF86SjCX+0h7IyMpwX/+c/B0f3RKaDOHgwOLxzzpyKdYQeB9GFv21bE37DMBKA4mL3+sUvKu7zhL+wMLgtck7cE04IL3vCH02gPXEPDfFURaTwt28P69ZVPM4T+Z/8BAYNcutZWeHHfPWVe5IYoFcvV0+0cFR1wl9W5l6HEib8hmHUCE/Uoo3C8YT/66+D27p3Dz+ma9fwcuvWbrljhxPqmTOD+yqLq1dGpPD36hWeLtkjNEzknRMp/F98AR9/7NYPHoRjj3U3Co+333Z1Rwr/JZfAww+7da9th9rwThN+wzBqRFWhC2+8vCf877wTTGIWyuzZ8OSTbr1bt6BAHnOMe+jKo6Ydo97oIY/QDt177gmuh4Z1POHv0iX83NWrg/9CPv/cLf/yFyfiqnDSSW5IaGTn7rPPuieBQ59ZiEw73diY8BuGUSOqEn5PBD3hj/T2PaZMgYsvduupqTB4sFtv2dL1CXgcfXTFc3/84/CnbqviqKNcPwGEh5yiCX/oTaZnT3j11WC/Q2iIZ+lSN0wV3BDSUI+/qCiYYrp58+B1DrU4vwm/YRg1IhbhLy52gu7lqq8OT/hTU8OFP3JcPsDzz7swTGVMnQoDB8LNNzsxX7DAjSAKrSs0hOR55aHiPnEiFBQEPf1QXnghOHw0LS0o/GVlwRsCBJ9ahkNP+G0cv2EYNSIW4QeXNjnWUE3fvm753Xfhwl8bvOGfHiJOoI88MvrxkYnlAEaPht/9rmKH8zHHwPz5wX8C7dq5ZxXatHEjmUJz8vt85vEbhpEgxCr848fHXmfv3m75zTdBL9mL+9cXoX0HoYQK/zvvwIoVFYeOgovZ5+aGZ/30/i20aeOWocIPwbaMHw8ffQT/+Y+b67exMeE3DKNGhIp75Ny1ofsmToy9Ti9Fsjfs8Z13YNWq2tlXGSkpcNZZFTtxx41zfQbXX+9CTvn5rq/B+4fgHd+ihUv6FoqX16cy4R8+3PVlrF0LL78Mp58O06dXnGqyoTHhNwyjRoR6/MOHh8e4Q5++7d8/9jozMuCuu+CNN1x58ODw7Jn1xYIF4R47uH6IL76APn3Ct993n1t6ydxUw4V/2rSg8HvPGXz7bXgdrVu70UutW4fH///737q1o66Y8Bu1InS2ISO5iAz1dO4MK1c6gdy/38X2Fy2qPulaJDfd5G4k8SQlJXa7LrzQZe288UZXPuMMlzbCI7TPwAtLffyxu4m99BI880xw/xFHhN8Uquqcbgisc9eoMQ8+CFdf7TyYyKcyjcQnUvh37YIBA4LlI4+smBe/qdKhg3sdOOA6iENDWaFPE8+YAUuWuNeoUS6kE8oRR8CyZcFyTVM/1zfm8Rs15tpr3fK99yru273beW2ffdawNhkNhyf8zZs7jzgyJBMZMkkEmjVz/xa8ztq0tGByueOOg3PPDYaG+vWreP4RR4SHmEz4jSbF/v3BSS3ef98tS0uD21audB1+//hH+Hk+HzzyiPMOjaaNJ/zr1ztveM0a5+lOmuQ6Ze+/v3Htizfbtrmc/d5Q1auvdssrr4Q//hFuvbXiOd4QVRGXpK6xhd9CPUaNWLMmuP6vf7kZj8aMca9nngkmxHrzzfDznnvO5Tn56COYMMF5S716xZ58y6ia225zN+C77or/tTzh98aot2wJw4a5VzLgZRedMsV1DJ99tiunplaevrlbN7dUdU8j79kTdzOrRLS62Q8OAQYOHKjLly9vbDMM4KmnXI7xu+5yT0aG8tFHbpwzuPHNu3a5sc8HD8J55znxj8TnOzQnqmhqRHsIKV7cdZfLrV9SUjE3jhGdXbtcaojJk92zCqrBEUzxRERWqOrAyO3m8cdIWZnLH/7QQ+5Of8wxjW1R47BqlfNsrrnG5WF5/30n3g88EBR9cOOUX3vNeTrZ2S55V7NmFTM6bt5ccVy1UTNCxb60NDiXbbzwPP5mzeJ7nUSiXTs31LVFC+cErV/fuPaYr4V7Um/LFjcE64QT3F83L6SxY4cbmtajh4tn3nYb3Htvo5rbIPznPy5Wu3Mn7NsX3P7xxy6BVbNmcP75rkNr5szgk5ehnH666+jzMjZ++qkbR/3II8FjQvO2G7UjtN8kWu75+qakxIV5ajpcM9lp1879Q2rbNklj/CJyGvAgkAo8pqp3N4YdK1e6DHw33RS+ffnyimGJrCwYO9YlaIo2U1CioOoE23vA5Jpr3E1v9Wo3gmPRIuexhNK2Lbz1Fpx6qnsK8rLL3D8AL33t1Knu30BamntCsrjYzXz08cdO+IcMiU87KhOmZcvcjf2kk9yIlIwMZ8emTa4Trls3196srNgnyi4shL//3T3B2bevuxlu2ODeuyuvdO9LvEJaoU+4fv558ClYVfdeh85NWx94wm/UjrZtGz/G3+DCLyKpwJ+AHwEbgPdF5HlVjZIHr26s2vIlazZuYM/eMkoOKKWlsGq1sn8/CCk8OSeVsoOp0LUZ+f1TKS7x0auXMHhQKi+9pKxaU0ZJUSrjxsGVV6TSonkqu68p4x/v76DP00p6uvsBbNnifgwlJXCgRCgtFXJyhMMPh/Q0IbOFsG0bfLOhjKISP82buUyBHQK9+1u2OE+gtBQOlARHAKSlu7/tzZuBX0H9bgmQIiAp7se9cWPwuC1b3bCx5s3gsE7uuO++E9q2dQKwdat72KRFC+eJex1Vu3bBypXCyvebsXWLQDcBfxqklLH2YCqHnXiAXTuBI1LonNueTzanoiiqiuKMevxFV9dW4KJrwz+LT7eFlx9ZCEN/uIu/vV3Cljau/aruPSgtdaG1lBR3s9i71+1v2RIyW0JmCxcy8voHUlPda+1adzNfudLFUY891mVRLC11Ig/KP55JZeV7zQGBlDLSUlNo01rYsT2lfBsoIBz7gwPMmOHn4EFh/z6hsBDWrxeys4X0dNiyWfjsMzhvovC3v7n9qLh6cOsbv4VX/ge4/gAXT4bMTB9HHy0uTKIpbNksZPcTUkXw+6HMr/h8is/np8znx++HomLw+1y1h3cSUtNg10733qSkwi9+kg7d0sGfxnk3FXHso9C6jfLddy6UdvJJsHMXpKZAZkvluOPcrFMpqQTukM5cEVBVWrYMjlk/WOrdOJTiYnf4R0WQ0j2TjzcHZz3x+gq974SIkCIppEoqLdLd3fNA2QHaZrRFVdlRvIOMtAxapLegzF/G3gN78av7a5iWkoZPfagqKZJCWkoafvXTtY2bwWXz/s2U+kppntacZqnNaNO8DTuLd9KyWUuKSosoKi3Cr35UleZpzVFVSsrCHz6QgFcgODsP+A6QlpJGWkoapb7S8u90LH2gIkJmeiZ+9eNXPz6/D5/6SE9xMbcyfxk+9ZUfv69tGbs67mT63UqXLm7oc1GRu0m3bOn+GXizlLVoAecMHsjhraKkKa0DDd65KyJDgJmqOiZQvhlAVSsdj1Dbzt1JP13P3Oe2gi+d0B+jM8QPKT4e+EMZOf0PkJqmpEpq+YcnIqRKKj71IQhl/jL86mfOk36e/EtH8EULcAbeS9Ho6/400EMguiaVfObio13HMn7Qs4wOh/kYepKfjz5M5eijlTcXZ/DRR8Ix3Q9wx9176HiYv/xHI7X8z3/JRc3Z+HX74GdSE1srIS2timnuVMjILOP8iw7QvoPy3jvpHNbJz+YtSnY/P80zlFWfpZKWKqgqi15Jd9+dyj7PiPXbboOcHGXDd8rRR0NKitKxI9zxG+W1V5q5z96fGmyT+N35om7dex80Jfiq9v1QSD3IyaOKGTTEx9zZbWiRIZSWOtHYtUvYvMmNJDlwEIqLYO/eiPe7wvtf3X7o1mcbjz4ePq2UBM7zvg+eAPr8vgrnp6akhm1PT00nVZwzUeorJS0ljYO+YIeQiMQkwl7dKeLeu1JfaVj9QLmoezaW+ctokd4Cv/op85eRlpJWfr5EvhdR8Ku//MaSmpJKqrjrl/pLEYS0lDRSU4J5oF9+Ge77bXvwuxtDSsCBO/JI5zxu3x5e/1MPdeOiibXLWFdZ525jCP8E4DRVnR4oTwYGqerPI46bAcwAOOaYYwasr0VvyKJF7q9vmzbuRyDiPJ0OHdybu3t3eIdkLOzb58IaqanOU2/WzHVetmwZfJWWuo7NvXudZ+p5rP36ueFf+/a5UENRkfvbd9xxzrsqLXVfgr17nTd+8KDbVlISfNTce6m6Y/z+YEfzvn1utiNv6Nj27c4rbtfOefrgvlzFxW48fkqKC1Gkp7sRB4cfHn3ii3ixc6f7fHbudDalpLj3qXlzZ5PP59rfrp3bv2eP+8z27HGfp3eMN6dply7usfqdO91nnJbmxlynpsLrr7t6+vaNvTP5k0/ce5iR4T7XHj3c+o4d7rPZt899bh984OyYPDn6VIFlZW5iksxMZ/tXX7l6vFEx330X/NeSlhZc914ZGe46Iu6f2cGD7ntUVuY+xz59Kv8e+/3u+xQ6bPajj4ITpXg//1AZ2LbNlVu0cPZ5YYnWrd0109Pd0M1Yc+mUlJWUOwmb9m2iqLSI4zsez0HfQXx+H4rSqlmrCuftLtlNqb+UVs1akSqpbNizgTJ/Gd3adaPMX8b2ou20zWjLrpJdtM9oz86SnfjVT1a7rPI6DvoOsmX/Frq06hImvh6qygHfATLS6ha78qu//GZRHaru9+jzuX+m2dnuM/a+Oxs2uHBks2buu3bmmdHnJYiFQ0n4JwJjIoT/RFWNMnWzw4ZzGoZh1JzKhL8x4g4bgFC/sisQZTpkwzAMIx40hvC/DxwvIt1FpBlwAfB8I9hhGIaRlDT4qB5VLRORnwP/xQ3nfEJVLaWXYRhGA9Eo4/hV9SXgpca4tmEYRrJzCIwtNAzDMBoSE37DMIwkw4TfMAwjyTDhNwzDSDKaRD5+EdkK1DaR6WHAtmqPShySrb2QfG229iY+9dXmbqraKXJjkxD+uiAiy6M9uZaoJFt7IfnabO1NfOLdZgv1GIZhJBkm/IZhGElGMgj/o41tQAOTbO2F5GuztTfxiWubEz7GbxiGYYSTDB6/YRiGEYIJv2EYRpKR0MIvIqeJyBoRWSsiN1V/xqGPiDwhIltE5NOQbR1E5FUR+TKwbB+y7+ZA+9eIyJjGsbr2iMjRIlIgIqtE5DMRuSqwPSHbLCIZIvKeiHwUaO8dge0J2V4PEUkVkQ9E5IVAOdHbWygin4jIhyKyPLCt4dqsqgn5wqV8/go4FmgGfAT0aWy76qFdw4F84NOQbfcANwXWbwL+X2C9T6DdzYHugfcjtbHbUMP2dgHyA+utgS8C7UrINuMmvG0VWE8HlgGDE7W9Ie2+Bvg78EKgnOjtLQQOi9jWYG1OZI//RGCtqn6tqgeBecC4RrapzqjqEmBHxOZxwJzA+hzg7JDt81T1gKquA9bi3pcmg6puVNWVgfW9wCrgKBK0zerYFyimB15KgrYXQES6AmcCj4VsTtj2VkGDtTmRhf8o4NuQ8obAtkSks6puBCeUwOGB7Qn1HohIFtAf5wUnbJsDYY8PgS3Aq6qa0O0FHgBuAPwh2xK5veBu5q+IyAoRmRHY1mBtbpSJWBoIibIt2cauJsx7ICKtgGeBq1V1j0i0prlDo2xrUm1WVR+QJyLtgAUikl3F4U26vSIyFtiiqitEZGQsp0TZ1mTaG8JJqvq9iBwOvCoiq6s4tt7bnMgefzJN6r5ZRLoABJZbAtsT4j0QkXSc6D+tqv8KbE7oNgOo6i5gMXAaidvek4CzRKQQF479oYg8ReK2FwBV/T6w3AIswIVuGqzNiSz8yTSp+/PAlMD6FOC5kO0XiEhzEekOHA+81wj21Rpxrv3jwCpV/X3IroRss4h0Cnj6iEgL4FRgNQnaXlW9WVW7qmoW7jf6uqpeTIK2F0BEWopIa28dGA18SkO2ubF7t+Pcc34GbhTIV8CtjW1PPbVpLrARKMV5ApcBHYHXgC8Dyw4hx98aaP8a4PTGtr8W7T0Z97f2Y+DDwOuMRG0zkAN8EGjvp8Btge0J2d6Ito8kOKonYduLG2n4UeD1madNDdlmS9lgGIaRZCRyqMcwDMOIggm/YRhGkmHCbxiGkWSY8BuGYSQZJvyGYRhJRiI/uWsYNUZEfMAnuN/GOmCyugepDCNhMI/fMMIpVtU8Vc3GJcP7WWMbZBj1jQm/YVTOOwSSYYlInoi8KyIfi8gCEWkvIoeLyIrA/lwRURE5JlD+SkQyRWSiiHwayK+/pBHbYhjlmPAbRhREJBU4hWCaj78BN6pqDi4UdLu6PCsZItIGGAYsB4aJSDdc4rEi4DZgjKrmAmc1dDsMIxom/IYRTotASuTtQAdc5sS2QDtVfSNwzBzchDgAb+MSjQ0H/jewHAYsDex/C5gtIpfjJgcyjEbHhN8wwilW1TygG27mtupi/EtxQt8Nl1QrF5dfaAmAql4J/AqXXfFDEekYH7MNI3ZM+A0jCqq6G/glcB1QBOwUkWGB3ZMBz/tfAlwMfKmqflyH8Bk4Tx8ROU5Vl6nqbcA2wtPrGkajYMM5DaMSVPUDEfkIly54CvCIiGQCXwPTAscUBiaF8Tpu3wS6qurOQPleETkeN5nGa7iMjIbRqFh2TsMwjCTDQj2GYRhJhgm/YRhGkmHCbxiGkWSY8BuGYSQZJvyGYRhJhgm/YRhGkmHCbxiGkWT8fxV78fpa0B57AAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "a = df.loc[df['CRIM_anomalies'] == -1] #anomaly\n",
    "plt.figure(figsize=(6,4))\n",
    "plt.plot(df['CRIM'], color='blue', label='original CRIM ')\n",
    "plt.plot(a['CRIM'], linestyle='none', marker='X', color='red', markersize=8, label='Anomaly - cities with the highest crime rate')\n",
    "plt.plot(df['decision_function'], color='green', label='decision function', alpha=0.3) # The anomaly score of the input samples.\n",
    "#             The lower, the more abnormal. Negative scores represent outliers,\n",
    "#             positive scores represent inliers.\n",
    "plt.xlabel('Rows')\n",
    "plt.ylabel('CRIM value')\n",
    "plt.title('Anomalies')\n",
    "plt.legend(loc='best')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "4777da29",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "380    88.9762\n",
       "404    41.5292\n",
       "405    67.9208\n",
       "410    51.1358\n",
       "414    45.7461\n",
       "418    73.5341\n",
       "Name: CRIM, dtype: float64"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# cities with the highest crime rate\n",
    "a['CRIM']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "66c1ce66",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#now we try ISO FOREST for different outlier fraction and different target feature\n",
    "# and we create fake outlier to the unrealistically high tax \n",
    "warnings.filterwarnings(\"ignore\")\n",
    "df['TAX'][0]=3000\n",
    "\n",
    "outliers_fraction=0.01\n",
    "df = iso_forest_detection(df, outliers_fraction, 'TAX', plotting=True)"
   ]
  }
 ],
 "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": "3.8.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}