score_card.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import matplotlib.cm as cm\n",
    "import seaborn as sns\n",
    "%matplotlib inline\n",
    "\n",
    "from sklearn import preprocessing\n",
    "from scipy.stats import skew, boxcox\n",
    "import zipfile\n",
    "import os"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "dataset_path = '/home/airflow/data/gao/data/'\n",
    "zipfile_path = os.path.join(dataset_path, 'paysim1.zip')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "csvfile_path = os.path.join(dataset_path, 'PS_20174392719_1491204439457_log.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "raw_data = pd.read_csv(csvfile_path)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "   step      type    amount     nameOrig  oldbalanceOrg  newbalanceOrig  \\\n",
      "0     1   PAYMENT   9839.64  C1231006815       170136.0       160296.36   \n",
      "1     1   PAYMENT   1864.28  C1666544295        21249.0        19384.72   \n",
      "2     1  TRANSFER    181.00  C1305486145          181.0            0.00   \n",
      "3     1  CASH_OUT    181.00   C840083671          181.0            0.00   \n",
      "4     1   PAYMENT  11668.14  C2048537720        41554.0        29885.86   \n",
      "\n",
      "      nameDest  oldbalanceDest  newbalanceDest  isFraud  isFlaggedFraud  \n",
      "0  M1979787155             0.0             0.0        0               0  \n",
      "1  M2044282225             0.0             0.0        0               0  \n",
      "2   C553264065             0.0             0.0        1               0  \n",
      "3    C38997010         21182.0             0.0        1               0  \n",
      "4  M1230701703             0.0             0.0        0               0  \n",
      "数据统计信息：\n",
      "               step        amount  oldbalanceOrg  newbalanceOrig  \\\n",
      "count  6.362620e+06  6.362620e+06   6.362620e+06    6.362620e+06   \n",
      "mean   2.433972e+02  1.798619e+05   8.338831e+05    8.551137e+05   \n",
      "std    1.423320e+02  6.038582e+05   2.888243e+06    2.924049e+06   \n",
      "min    1.000000e+00  0.000000e+00   0.000000e+00    0.000000e+00   \n",
      "25%    1.560000e+02  1.338957e+04   0.000000e+00    0.000000e+00   \n",
      "50%    2.390000e+02  7.487194e+04   1.420800e+04    0.000000e+00   \n",
      "75%    3.350000e+02  2.087215e+05   1.073152e+05    1.442584e+05   \n",
      "max    7.430000e+02  9.244552e+07   5.958504e+07    4.958504e+07   \n",
      "\n",
      "       oldbalanceDest  newbalanceDest       isFraud  isFlaggedFraud  \n",
      "count    6.362620e+06    6.362620e+06  6.362620e+06    6.362620e+06  \n",
      "mean     1.100702e+06    1.224996e+06  1.290820e-03    2.514687e-06  \n",
      "std      3.399180e+06    3.674129e+06  3.590480e-02    1.585775e-03  \n",
      "min      0.000000e+00    0.000000e+00  0.000000e+00    0.000000e+00  \n",
      "25%      0.000000e+00    0.000000e+00  0.000000e+00    0.000000e+00  \n",
      "50%      1.327057e+05    2.146614e+05  0.000000e+00    0.000000e+00  \n",
      "75%      9.430367e+05    1.111909e+06  0.000000e+00    0.000000e+00  \n",
      "max      3.560159e+08    3.561793e+08  1.000000e+00    1.000000e+00  \n",
      "数据集基本信息：\n",
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 6362620 entries, 0 to 6362619\n",
      "Data columns (total 11 columns):\n",
      "step              int64\n",
      "type              object\n",
      "amount            float64\n",
      "nameOrig          object\n",
      "oldbalanceOrg     float64\n",
      "newbalanceOrig    float64\n",
      "nameDest          object\n",
      "oldbalanceDest    float64\n",
      "newbalanceDest    float64\n",
      "isFraud           int64\n",
      "isFlaggedFraud    int64\n",
      "dtypes: float64(5), int64(3), object(3)\n",
      "memory usage: 534.0+ MB\n",
      "None\n"
     ]
    }
   ],
   "source": [
    "print(raw_data.head())\n",
    "\n",
    "print('数据统计信息：')\n",
    "print(raw_data.describe())\n",
    "\n",
    "print('数据集基本信息：')\n",
    "print(raw_data.info())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "转账类型记录统计：\n",
      "CASH_OUT    2237500\n",
      "PAYMENT     2151495\n",
      "CASH_IN     1399284\n",
      "TRANSFER     532909\n",
      "DEBIT         41432\n",
      "Name: type, dtype: int64\n"
     ]
    }
   ],
   "source": [
    "\n",
    "# 查看转账类型\n",
    "print('转账类型记录统计：')\n",
    "print(raw_data['type'].value_counts())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAgMAAAE0CAYAAACxTDDfAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nDEiS1HOGAUmSes4wIElSz/1/RsVb2fw2pmMAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fec5b346be0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(1, 1, figsize=(8, 4))\n",
    "raw_data['type'].value_counts().plot(kind='bar', title='Transaction Type', ax=ax, figsize=(8, 4))\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Figure(576x288) Axes(0.125,0.125;0.775x0.755)\n"
     ]
    }
   ],
   "source": [
    "print(fig, ax)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Rectangle(-0.25,0;0.5x1.39928e+06)\n",
      "Rectangle(0.75,0;0.5x2.23338e+06)\n",
      "Rectangle(1.75,0;0.5x4116)\n",
      "Rectangle(2.75,0;0.5x41432)\n",
      "Rectangle(3.75,0;0.5x2.1515e+06)\n",
      "Rectangle(4.75,0;0.5x528812)\n",
      "Rectangle(5.75,0;0.5x4097)\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAaEAAAFWCAYAAADAPyJjAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAIABJREFUeJzs3Xl8VNX9//HXm00UpIASyyKLiLIlrFUq1rVsBbTgSlWo\nS3Gjghui31/rVhW1IqXgLhVBpVVRUVSKgNalikERELGgBgkimyD7/vn9cW/iJGSSSUhyZ8jn+XjM\nI3fO3T4zuTOfueeee47MDOeccy4KlaIOwDnnXMXlScg551xkPAk555yLjCch55xzkfEk5JxzLjKe\nhJxzzkXGk5ArE5KOkPQfSZskPRB1PGVF0iOS/hR1HCUh6XJJo6OOo6Qk3SLpiajjKIykpyT9JZzO\nkPRB1DElG09CLg9JcyQdI+koSZ/sx6YGA2uBWmZ2fQH7yf1wpgpJv5f0XmyZmV1hZndGFVNJSaoG\n/D/g/vB5U0kmqUq0kSXOzO42s8uKWk7S25K2S9oc8/hlecQYy8zmAxsk9S3vfSczT0Iul6SqQBNg\nCdAJ2J8k1ARYZCW8GzqVvgxT1JnAYjNbEXUg+Ukqizvoh5hZzZjHfwvYb3kcc88Al5fDflKGJyEX\nqy0/JY7OFJGEJJ0g6WNJP4Z/TwjLnwIGAcPDX52/zrfeYOCCmPmvhuVZkm6SNB/YIqmKpBGSvgqr\n9RZJ6heznd9Lek/SXyWtl/SNpF755n8drvuNpAvC8uaSZklaJ2mtpGck1Y5Z70hJUyStCZcZK6kV\n8AjwyzDmDTmvNfaMTtIfJC2V9IOkqZIaxMwzSVdIWiJpg6RxkhTOO1rSO+F7uVbSP+O8529IGpKv\n7DNJ/RV4UNJqSRslLZDUNs6/rxfwTszz/4R/N4Sv7+TwNaTH7CdN0lZJ9SSdIik7rBJbG/7vLohZ\n9qDw//KtpFVhteXBcWIpEUm3SZoUTleXNCn8f20Ij8cjEtiGSbpa0hKCH19I+puk5eF7OFfSr2KW\nz///PkVSdszzDpI+CY+5fwLV8+3ybeB0SQft14s/kJiZPyr4A7gY2ABsBbaH07uBTeF0swLWqQus\nBy4CqgADwueHhfOfAv5SyD73mQ9kAfOAI4GDw7JzgAYEP5jOA7YA9cN5vwd2AX8AKgNXAt8BAmoA\nG4Fjw2XrA23C6aOBbsBBQD2CL+DR4bzKwGfAg+E2qgMnxuzvvXivAziNoAqyY7jtvwP/iVnWgNeA\n2kBjYA3QM5z3HPB/4evM3WcB79tA4P2Y563D/9FBQA9gbrh9Aa1y3qsCtvMxcE7M86ZhfFViyh4C\n7o15PhR4NZw+JTxGRoX7Pjn83+S83w8CU8Pj5FDgVeCeBI9HS3C524BJ4fTl4T4OCf+HnQiqgiH4\n4r8s3r6AGWGcOcfchcBhBMf19cD3QPWCjtvwfcgOp6sBy4BrgarA2QTHZ/7jfCOQEfXnPlkefibk\nMLN/mFltgi+wLkAGsJDgQ1zbzL4pYLXewBIzm2hmu83sOWAxsL/13WPMbLmZbQtje97MvjOzvWb2\nT4Jfq8fFLL/MzB43sz3ABIJkk/MLeC/QVtLBZrbSzD4Pt7nUzGaY2Q4zW0PwRXpyuM5xBEnvRjPb\nYmbbzSzPdaBCXACMN7NPzGwHcDPBmVPTmGVGmtkGM/sWmA20D8t3EVRhNihiny8B7SU1idnnlHB/\nuwi+8FsCMrMvzGxlnO3UJviRUZgJwICcszWCHxwT8y3zp/B9fAeYBpwbLj8YuNbMfjCzTcDdwPlF\n7G9/7CJIHEeb2R4zm2tmG2PmjwnPkDZo32ud94Rx5hxzk8xsXXhcP0CQZI9NIIYuBMlntJntMrMX\nCJJ9fpsI3n+HV8dVeJLqhh/MH4ETCH41fknwoVsvaVicVRsQ/OqLtQxouJ8hLc8X30BJ83K+QAiq\nDA+PWeT7nAkz2xpO1jSzLQRnTlcAKyVNk9Qy3OYRkiZLWiFpIzApZptHEiS23SWIPc97YmabgXXk\nfU++j5neCtQMp4cTnL3MkfS5pEsK2kH4hT6Nn77QBxBcZ8DMZgFjgXHAakmPSaoVJ9b1BAkrLjP7\nKIzxlPC9O5rg7CZ3G+H7nGMZwXtQj+CMZG7M/+3NsHwfkk6MSRA51ZwbYh4nFhZnaCIwHZgs6TtJ\n9ym4xpnjmvAHVW0z65hv3fzH3A2SvgirRjcAPyPvMRdPA2CFhac7ofyfEQje9w0JbK9C8CRUwYW/\nAGsTVGc8EU6/CfQNP7DxmvB+R/DLPVZjINEL3fEuPueWh7/2HweGEFTz1SY4Q1OcdfNuyGy6mXUj\nODtaHG4Lgl/lBqSbWS2C6pecbS4HGqvgi9RFXTDP855IqkHw67zI98TMvjezP5hZA4L/xUOSjo6z\n+HMEZyi/JKi6mx2znTFm1omgmu4Y4MY425gfzs9dNc5yEwjen4uAF8xse8y8OuFrzNGY4D1YC2wj\nqP7M+eL/mZnVpABm9l7McrXDstoxjyLPRMMzj9vNrDXBj6k+BFWXiYg95n5F8IPgXKBOGM+P/HR8\nbCFIsDl+HjO9EmgYc+YIwXuSS1JDgmq7LxOM7YDnScjliG0N14Ggaq4wrwPHSPqdggYE5xF88b2W\n4P5WAUcVsUwNgi+INQCSLiY4EypSeLZzZvgluQPYTFA9B8Ev0c3Aj+GXQuwX9RyCL5ORkmqEF7y7\nxsTcSEHz5oI8B1wsqX144flu4CMzy0og3nMkNQqfrg9f9944i79OkOzuAP5pZnvDbfxC0vHhGcAW\ngut7hW3j5Jjna8Jl8/9PJgH9CBLR0wVs53ZJ1cIv7z7A82E8jwMPSkoLY2soqUecWPabpFMlpUuq\nTHDNZRfxX3thDiW41rUGqCLpz0Ds2eQ84DdhDcLPgdiagv+G614jqaqk/uStOobgPZ8VVp86PAm5\nn3QCPpF0GLDHzNYXtrCZrSP40rmeoMppONDHzNYmuL8ngdZhdcvLcfaxCHiA4MO9CkgH3k9w+5WA\n6wh+mf9A8OG/Mpx3O0HjgR8JqramxOxzD8F1raOBb4Fsgmo9gFnA58D3kvZ5nWb2FvAn4EWCRNac\nxK+D/AL4SNJmgiqvoWb2dUELhl9gU4BfA8/GzKpF8OW/nqAaaB3hfUAFeBVoqbD1XliVeRfwfvg/\n6RKWLyf4cWLAu/m28X24r+8IqgSvMLPF4bybgKXAh2GV51skdl2lpH4OvECQgL4gaPmX//pVIqYT\n1AT8j+A93E7e6rqJBA1XsoB/A7mtGM1sJ9CfoAHLDwTHzRTyuoCglaULKW/1pXOuolDQVL61mcW7\n7pez3HjgOzP7fzFlpxC0TGsUd0WXh6QM4FEzK/cbZZOZJyHnXFxhy755QIfYVpKehFxp8eo451yB\nJN1J0BDk/jjN9J3bb34m5JxzLjJ+JuSccy4y3klkEQ4//HBr2rRp1GE451xKmTt37lozK/AG5Vie\nhIrQtGlTMjMzow6jXC1fvpyBAweyatUqJDF48GCGDh2aZ5lRo0bxxBNPUKVKFerVq8f48eNp0qQJ\ny5Yto1+/fuzdu5ddu3bxxz/+kSuuuAKAnj17snLlSnbv3s2vfvUrxo0bR+XKlfNs95FHHsktr1mz\nJo899hitW7dm165dXHbZZXzyySfs3r2bgQMHcvPNN+eut2fPHjp37kzDhg157bVEb1VyzpUVSQX1\nFrGvojqXq+iPTp06WUXz3Xff2dy5c83MbOPGjdaiRQv7/PPP8ywza9Ys27Jli5mZPfTQQ3buueea\nmdmOHTts+/btZma2adMma9Kkia1YscLMzH788UczM9u7d6/179/fnnvuuX32nbOMmdkrr7xiPXr0\nMDOzZ555xs477zwzM9uyZYs1adLEvvnmm9xlH3jgARswYID17t17v1+/c27/AZnmHZi6kqhfvz4d\nOwbdax166KG0atWKFSvy9jxz6qmncsghQe8lXbp0ITs76M2+WrVqHHRQ0Ev9jh072Lv3p5vWa9UK\nbjzfvXs3O3fuJG/vJnmXAdiyZUvuMpLYsmULu3fvZtu2bVSrVi132ezsbKZNm8ZllxU5vplzLsl4\nEnKFysrK4tNPP+X444+Pu8yTTz5Jr165w/iwfPlyMjIyOPLII7npppto0CB3SB169OhBWloahx56\nKGeffXaB2xs3bhzNmzdn+PDhjBkzBoCzzz6bGjVqUL9+fRo3bswNN9xA3bp1ARg2bBj33XcflSr5\n4excqvFPrYtr8+bNnHXWWYwePTrPGUqsSZMmkZmZyY03/tT92pFHHsn8+fNZunQpEyZMYNWqVbnz\npk+fzsqVK9mxYwezZs0qcJtXX301X331Fffeey9/+UswfticOXOoXLky3333Hd988w0PPPAAX3/9\nNa+99hppaWl06tSpFF+5c668eBJyBdq1axdnnXUWF1xwAf379y9wmbfeeou77rqLqVOn5lbBxWrQ\noAFt27bl3XfzdjlWvXp1zjzzTF555ZVCYzj//PN5+eWgW7lnn32Wnj17UrVqVdLS0ujatSuZmZm8\n//77TJ06laZNm3L++ecza9YsLrzwwhK+audcefMk5PZhZlx66aW0atWK6667Lrd87NixjB07FoBP\nP/2Uyy+/nKlTp5KWlpa7THZ2Ntu2bQNg/fr1vPfeexx77LFs3ryZlSuD8dV2797NtGnTaNmy5T7b\nXbJkSe62pk2bRosWLQBo3Lhx7pnTli1b+PDDD2nZsiX33HMP2dnZZGVlMXnyZE477TQmTZpUVm+N\nc66UeRNtt4/333+fiRMnkp6eTvv2wcCfd999N4sXL6Zr12BUgxtvvJHNmzdzzjnnAEGSmDp1Kl98\n8QXXX389kjAzbrjhBtLT01m1ahVnnHFGbmOFU089Nbfpdux2x44dy1tvvUXVqlWpU6cOEyZMAIIq\nuosvvpg2bdpgZlx88cVkZGSU91vjnCttiTShq8iPithEO57evXvbjh07Uma7rvx9++23dsopp1ir\nVq2sdevWNnr06H2Weeedd6xDhw5WuXJle/755/PMq1SpkrVr187atWtnffv2zS3/+9//bs2bNzfA\n1qxZU2gMy5Ytsxo1atj999+fWzZ69Ghr06aNtW7d2h588MHc8ltvvdUaNGiQu89p06aV9KW7fEiw\nibafCbmEldVNoH5z6YGjSpUqPPDAA3Ts2JFNmzbRqVMnunXrRuvWrXOXady4MU899RR//etf91n/\n4IMPZt68efuUd+3alT59+nDKKacUGcN1112Xp7XmwoULefzxx5kzZw7VqlWjZ8+e9OnTh6OPDgau\nvfbaa7nhhhtK8GpdafAk5PJoOmJamW4/a2TvMt2+i1b9+vWpX78+kPces9gklNMNVnGa1Hfo0CGh\n5V5++WWaNWtGjRo/jTr+xRdfcPzxx+fe13byySczZcoUhg8fnvD+XdnxhgnOuTKRyD1m+W3fvp3O\nnTvTpUuX3JaRidq8eTP33nsvt956a57ynBaa69atY+vWrbz++ussX/7TYKljx44lIyODSy65hPXr\nCx1Q2JUBT0LOuVKXyD1mBVm2bBmZmZk8++yzDBs2jK+++irhdW+77TauvfZaatasmae8VatW3HTT\nTXTv3p2ePXvSvn373D4Lr7zySr766ivmzZtH/fr1uf766xPenysdXh3nnCtVidxjFk/Dhg0BOOqo\nozjllFP49NNPad68eULrfvTRR7zwwgsMHz6cDRs2UKlSJapXr86QIUO49NJLufTSSwG45ZZbaNQo\nGBD2iCOOyF3/D3/4A3369ClWvG7/+ZmQc67UWAL3mMWzfv16duzYAcDatWt5//3381xLKsicOXMY\nOHAgAO+++y5ZWVlkZWUxbNgwbrnlFoYMGQLA6tWrAfj222+ZMmUKv/vd7wBy710DeOmll2jbtm0x\nX7HbX56EnHOlJuces1mzZtG+fXvat2/P66+/zuLFiznssMMA+Pjjj2nUqBHPP/88l19+OW3atAGC\nBgSdO3emXbt2nHrqqYwYMSI3CY0ZM4ZGjRqRnZ1NRkZGbme13377LQcffHCRcZ111lm0bt2avn37\nMm7cOGrXrg3A8OHDSU9PJyMjg9mzZ/Pggw+WxdviCuHDexehc+fOVpHGE/LWca4s9OnThylTplCt\nWrVS3e6NN97IRRdd5DcuJyFJc82sc1HL+TUh51ypKPQHTNsrOebPM/Zr+wX9gLn//vv3a5suel4d\n55xzLjKehJxzzkXGk5BzzrnIeBJyzjkXGU9CzjnnIuNJyDnnXGQ8CTnnnIuMJyHnnHOR8STknHMu\nMp6EnHPORabMkpCkIyXNlrRI0ueShobldSXNkLQk/FsnLJekMZKWSpovqWPMtgaFyy+RNCimvJOk\nBeE6YySppPtwzjlX/sryTGg3cL2ZtQa6AFdLag2MAGaaWQtgZvgcoBfQInwMBh6GIKEAtwLHA8cB\nt+YklXCZP8Ss1zMsL9Y+nHPORaPMkpCZrTSzT8LpTcAXQEPgTGBCuNgE4Lfh9JnA0xb4EKgtqT7Q\nA5hhZj+Y2XpgBtAznFfLzD60oCvwp/Ntqzj7cM45F4FyuSYkqSnQAfgIOMLMckaS+h7IGdqwIbA8\nZrXssKyw8uwCyinBPvLHO1hSpqTMNWvWJPYinXPOFVuZJyFJNYEXgWFmtjF2XngGU6YDGpVkH2b2\nmJl1NrPO9erVK6PInHPOlWkSklSVIAE9Y2ZTwuJVOVVg4d/VYfkK4MiY1RuFZYWVNyqgvCT7cM45\nF4GybB0n4EngCzMbFTNrKpDTwm0Q8EpM+cCwBVsX4MewSm060F1SnbBBQndgejhvo6Qu4b4G5ttW\ncfbhnHMuAmU5smpX4CJggaR5YdktwEjgX5IuBZYB54bzXgd+AywFtgIXA5jZD5LuBD4Ol7vDzH4I\np68CngIOBt4IHxR3H84556JRZknIzN4DFGf26QUsb8DVcbY1HhhfQHkm0LaA8nXF3Ydzzrny5z0m\nOOeci4wnIeecc5HxJOSccy4ynoScc85FxpOQc865yHgScs45FxlPQs455yLjScg551xkPAk555yL\njCch55xzkfEk5JxzLjKehJxzzkXGk5BzzrnIeBJyzjkXGU9CzjnnIuNJyDnnXGQ8CTnnnIuMJyHn\nnHOR8STknHMuMp6EnHPORcaTkHPOuch4EnLOORcZT0LOOeci40nIOedcZDwJOeeci4wnIeecc5Hx\nJOSccy4ynoScc85FxpOQc865yHgScs45FxlPQs455yLjScg551xkPAk555yLTJWiFpDUFbgNaBIu\nL8DM7KiyDc0559yBrsgkBDwJXAvMBfaUbTjOOecqkkSS0I9m9kaZR+Kcc67CSSQJzZZ0PzAF2JFT\naGaflFlUzjnnKoREktDx4d/OMWUGnFb64TjnnKtIimwdZ2anFvAoMgFJGi9ptaSFMWW3SVohaV74\n+E3MvJslLZX0paQeMeU9w7KlkkbElDeT9FFY/k9J1cLyg8LnS8P5TYvah3POuWgUmYQk/UzSKEmZ\n4eMBST9LYNtPAT0LKH/QzNqHj9fDfbQGzgfahOs8JKmypMrAOKAX0BoYEC4LcG+4raOB9cClYfml\nwPqw/MFwubj7SOB1OOecKyOJ3Cc0HtgEnBs+NgL/KGolM/sP8EOCcZwJTDazHWb2DbAUOC58LDWz\nr81sJzAZOFOSCKoDXwjXnwD8NmZbE8LpF4DTw+Xj7cM551xEEklCzc3s1jARfG1mtwP7c4/QEEnz\nw+q6OmFZQ2B5zDLZYVm88sOADWa2O195nm2F838Ml4+3rX1IGpxz5rdmzZqSvUrnnHNFSiQJbZN0\nYs6T8ObVbSXc38NAc6A9sBJ4oITbKVNm9piZdTazzvXq1Sv2+pdccglpaWm0bdu2wPmPPPII6enp\ntG/fnhNPPJFFixYBsHPnTi6++GLS09Np164db7/9du46//znP8nIyKBNmzbcdNNNBW73mWeeISMj\ng/T0dE444QQ+++yz3HkPPvggbdq0oW3btgwYMIDt27fnWfeaa66hZs2axX6tzjm3PxJJQlcC4yRl\nSVoGjAWuKMnOzGyVme0xs73A4/xUHbYCODJm0UZhWbzydUBtSVXylefZVjj/Z+Hy8bZV6n7/+9/z\n5ptvxp3/u9/9jgULFjBv3jyGDx/OddddB8Djjz8OwIIFC5gxYwbXX389e/fuZd26ddx4443MnDmT\nzz//nO+//56ZM2fus91mzZrxzjvvsGDBAv70pz8xePBgAFasWMGYMWPIzMxk4cKF7Nmzh8mTJ+eu\nl5mZyfr160vzLXDOuYQk0jpunpm1AzKAdDPrYGafFbVeQSTVj3naD8hpOTcVOD9s2dYMaAHMAT4G\nWoQt4aoRNCyYamYGzAbODtcfBLwSs61B4fTZwKxw+Xj7KHUnnXQSdevWjTu/Vq1audNbtmwhuGQF\nixYt4rTTgoaHaWlp1K5dm8zMTL7++mtatGhBzlnZr3/9a1588cV9tnvCCSdQp05Qw9mlSxeys7Nz\n5+3evZtt27axe/dutm7dSoMGDQDYs2cPN954I/fdd99+vmrnnCu+uPcJSbrQzCZJui5fOQBmNqqw\nDUt6DjgFOFxSNnArcIqk9gT3GWUBl4fb+lzSv4BFwG7gajPbE25nCDAdqAyMN7PPw13cBEyW9Bfg\nU4LuhQj/TpS0lKBhxPlF7SMK48aNY9SoUezcuZNZs2YB0K5dO6ZOncqAAQNYvnw5c+fOZfny5Zx2\n2ml8+eWXZGVl0ahRI15++WV27txZ6PaffPJJevXqBUDDhg254YYbaNy4MQcffDDdu3ene/fuAIwd\nO5YzzjiD+vXrF7Y555wrE4XdrFoj/HtoAfOsqA2b2YACip8soCxn+buAuwoofx14vYDyrymgdZuZ\nbQfOKc4+onD11Vdz9dVX8+yzz/KXv/yFCRMmcMkll/DFF1/QuXNnmjRpwgknnEDlypWpU6cODz/8\nMOeddx6VKlXihBNO4Kuvvoq77dmzZ/Pkk0/y3nvvAbB+/XpeeeUVvvnmG2rXrs0555zDpEmTOO20\n03j++efzXHtyzrnyFDcJmdmj4eRbZvZ+7LywcYIrBeeffz5XXnklAFWqVOHBBx/MnXfCCSdwzDHH\nANC3b1/69u0LwGOPPUblygXf4jR//nwuu+wy3njjDQ477DAA3nrrLZo1a5Zbnde/f38++OAD6tSp\nw9KlSzn66KMB2Lp1Kyse/QMNL3+8bF6sc87lk0jDhL8nWOYKMXbsWMaOHQvAkiVLcsunTZtGixYt\ngCAJbNmyBYAZM2ZQpUoVWrcO7s1dvXo1EJzVPPTQQ1x22WUAvPTSS9x8880AfPvtt/Tv35+JEyfm\nJi+Axo0b8+GHH7J161bMjJkzZ9KqVSt69+7N999/T1ZWFllZWRxyyCGegJxz5aqwa0K/BE4A6uW7\nLlSL4PqMi2PAgAG8/fbbrF27lkaNGnH77bezePFiunYNTiDHjh3LW2+9RdWqValTpw4TJgT31q5e\nvZoePXpQqVIlGjZsyMSJE3O3OXTo0Nwm13/+859zk8xXX32V29DhjjvuYN26dVx11VVAcGaVmZnJ\n8ccfz9lnn03Hjh2pUqUKHTp0yG0555xzUVLQcKyAGdLJBA0LrgAeiZm1CXjVzJYUtN6BpnPnzpaZ\nmZnw8k1HTCuwfPULt1Ov3y2octX9iidrZO88zy+88EIefPBBSnI/U0HixV9a8sfvDhx+7LhYkuaa\nWeeilivsmtA7wDuSnjKzZaUaXQWUdvatZbLdSZMmlcl2nXOuPCRyTegJSbVznkiqI2l6GcbknHOu\ngkgkCR1uZhtynpjZeiCt7EJyzjlXUSSShPZKapzzRFITErhPyDnnnCtKIiOr/h/wnqR3AAG/Arxp\nlXPOuf1WZBIyszcldQS6hEXDzGxt2YblnHOuIkjkTAhgD7AaqA60lpQzaJ1zzjlXYkUmIUmXAUMJ\nhj6YR3BG9F+CkU2dc865EkukYcJQ4BfAMjM7FegAbCh8Feecc65oiSSh7WHP1Eg6yMwWA8eWbVjO\nOecqgkSuCWWHN6u+DMyQtB7wHhScc87tt0Rax/ULJ2+TNJtguOz4Y1c755xzCSqyOk5Sc0kH5TwF\nmgKHlGVQzjnnKoZErgm9COyRdDTwGHAk8GyZRuWcc65CSKjbHjPbDfQD/m5mNwL1yzYs55xzFUEi\nSWiXpAHAIOC1sGz/BsVxzjnnSCwJXQz8ErjLzL6R1AyYWMQ6zjnnXJESaR23CLgm5vk3wL1lGZRz\nzrmKIZFue7oCtwFNwuUFmJkdVbahOeecO9AlcrPqk8C1wFyCjkydc865UpFIEvrRzN4o80icc85V\nOIkkodmS7gemADtyCs3skzKLyjnnXIWQSBI6PvzbOabM8KEcnHPO7adEWsedWh6BOOecq3gSGllV\nUm+gDcHIqgCY2R1lFZRzzrmKIZEOTB8BzgP+SNA8+xyC5trOOefcfkmkx4QTzGwgsN7MbifoPeGY\nsg3LOedcRZDQyKrh362SGgC78A5MnXPOlYJErgm9Go6sej/wCUHLuMfLNCrnnHMVQqFJSFIlYKaZ\nbQBelPQaUN3MfiyX6Jxzzh3QCq2OM7O9wLiY5zs8ATnnnCstiVwTminpLEkq82icc85VKIkkocuB\n54EdkjZK2iRpYxnH5ZxzrgIoMgmZ2aFmVsnMqplZrfB5raLWkzRe0mpJC2PK6kqaIWlJ+LdOWC5J\nYyQtlTRfUseYdQaFyy+RNCimvJOkBeE6Y3LO1EqyD+ecc9FI5GbVmYmUFeApoGe+shEEDR1aADPD\n5wC9gBbhYzDwcLifusCtBP3XHQfcmpNUwmX+ELNez5LswznnXHTiJiFJ1cMkcLikOuEZRl1JTYGG\nRW3YzP4D/JCv+ExgQjg9AfhtTPnTFvgQqC2pPtADmGFmP5jZemAG0DOcV8vMPjQzA57Ot63i7MM5\n51xECmuifTkwDGhAMKBdTsOEjcDYEu7vCDNbGU5/DxwRTjcElscslx2WFVaeXUB5SfaxknwkDSY4\nW6Jx48YJvjTnnHPFFfdMyMz+ZmbNgBvM7CgzaxY+2plZSZNQ7PaN4MbXMlPSfZjZY2bW2cw616tX\nrwwic85gBxioAAAgAElEQVQ5B4k1TPh7Ke5vVU4VWPh3dVi+AjgyZrlGYVlh5Y0KKC/JPpxzzkUk\nkSbapWkqkNPCbRDwSkz5wLAFWxeCIcVXAtOB7uE1qTpAd2B6OG+jpC5hq7iB+bZVnH0455yLSNxr\nQpK6mtn7kg4ysx3xlitk/eeAUwgaNmQTtHIbCfxL0qXAMuDccPHXgd8AS4GtwMUAZvaDpDuBj8Pl\n7jCznMYOVxG0wDsYeCN8UNx9OOeci05hDRPGAJ2A/wLFvqfGzAbEmXV6AcsacHWc7YwHxhdQngm0\nLaB8XXH34ZxzLhqFJaFdkh4DGkoak3+mmV1TdmE555yrCApLQn2AXxPcqzO3fMJxzjlXkcRNQma2\nFpgs6Qsz+6wcY3LOOVdBJNI6bp2kl8J+4FZLelFSo6JXc8455wqXSBL6B0Hz5gbh49WwzDnnnNsv\niSShNDP7h5ntDh9PAd6NgHPOuf2WSBJaK+lCSZXDx4XAurIOzDnn3IEvkSR0CcENn98TdPZ5Nn6j\np3POuVJQWBNtAMxsGXBGOcTinHOuginvvuOcc865XJ6EnHPORcaTkHPOucgUmYQk/b+Y6YPKNhzn\nnHMVSdwkJOkmSb8kaA2X479lH5JzzrmKorDWcYuBc4CjJL0bPj9M0rFm9mW5ROecc+6AVlh13Abg\nFoJB4E4B/haWj5D0QRnH5ZxzrgIo7EyoB/BnoDkwCpgPbDEzv1HVOedcqYh7JmRmt5jZ6UAWMBGo\nDNST9J6kV8spPueccwewIntMAKaHQ2lnSrrSzE6UdHhZB+acc+7AV2QTbTMbHvP092HZ2rIKyDnn\nXMVRrJtVfYRV55xzpcl7THDOORcZT0LOOeci40nIOedcZDwJOeeci4wnIeecc5HxJOSccy4ynoSc\nc85FxpOQc865yHgScs45FxlPQs45F6Np06akp6fTvn17OnfuDMCNN95Iy5YtycjIoF+/fmzYsAGA\nXbt2MWjQINLT02nVqhX33HNPgdt87rnnSE9PJyMjg549e7J2bdDz2fPPP0+bNm2oVKkSmZmZucvP\nmDGDTp06kZ6eTqdOnZg1a1YZv+roeBJyzrl8Zs+ezbx583ITQ7du3Vi4cCHz58/nmGOOyU02zz//\nPDt27GDBggXMnTuXRx99lKysrDzb2r17N0OHDmX27NnMnz+fjIwMxo4dC0Dbtm2ZMmUKJ510Up51\nDj/8cF599VUWLFjAhAkTuOiii8r+RUfEk5BzzhWhe/fuVKkSDDrQpUsXsrOzAZDEli1b2L17N9u2\nbaNatWrUqlUrz7pmhpmxZcsWzIyNGzfSoEEDAFq1asWxxx67z/46dOiQu0ybNm3Ytm0bO3bsKMuX\nGBlPQs45F0MS3bt3p1OnTjz22GP7zB8/fjy9evUC4Oyzz6ZGjRrUr1+fxo0bc8MNN1C3bt08y1et\nWpWHH36Y9PR0GjRowKJFi7j00ksTjufFF1+kY8eOHHTQQfv3wpKUJyHnnIvx3nvv8cknn/DGG28w\nbtw4/vOf/+TOu+uuu6hSpQoXXHABAHPmzKFy5cp89913fPPNNzzwwAN8/fXXeba3a9cuHn74YT79\n9FO+++47MjIy4l47yu/zzz/npptu4tFHHy29F5hkPAk551yMhg0bApCWlka/fv2YM2cOAE899RSv\nvfYazzzzDJIAePbZZ+nZsydVq1YlLS2Nrl275mlgADBv3jwAmjdvjiTOPfdcPvjggyLjyM7Opl+/\nfjz99NM0b968NF9iUvEk5JxzoS1btrBp06bc6X//+9+0bduWN998k/vuu4+pU6dyyCGH5C7fuHHj\n3JZrW7Zs4cMPP6Rly5YAnH766axYsYKGDRuyaNEi1qxZAwQt31q1alVoHBs2bKB3796MHDmSrl27\nlsVLTRqRJCFJWZIWSJonKTMsqytphqQl4d86YbkkjZG0VNJ8SR1jtjMoXH6JpEEx5Z3C7S8N11Vh\n+3DOOYBVq1Zx4okn0q5dO4477jh69+5Nz549GTJkCJs2baJbt260b9+eK664AoCrr76azZs306ZN\nG37xi19w8cUXk5GRwd69e1m6dCl169alQYMG3HrrrZx00klkZGQwb948brnlFgBeeuklGjVqxH//\n+1969+5Njx49ABg7dixLly7ljjvuoH379rRv357Vq1dH9r6UJZlZ+e9UygI6xw4TLuk+4AczGylp\nBFDHzG6S9Bvgj8BvgOOBv5nZ8ZLqAplAZ8CAuUAnM1svaQ5wDfAR8DowxszeiLePwmLt3Lmz5T+9\nLkzTEdMSXrYkskb2LtPtp3r8LjqpfuyUZvw712Sxef4M6p7+h9yyinbsS5prZp2LWi6ZquPOBCaE\n0xOA38aUP22BD4HakuoDPYAZZvaDma0HZgA9w3m1zOxDCzLs0/m2VdA+nHOu1FSr1zRPAnLxRZWE\nDPi3pLmSBodlR5jZynD6e+CIcLohsDxm3eywrLDy7ALKC9tHHpIGS8qUlJlTj+ucc670VYlovyea\n2QpJacAMSYtjZ5qZSSrTesLC9mFmjwGPQVAdV5ZxOOdcRRbJmZCZrQj/rgZeAo4DVoVVaYR/c67C\nrQCOjFm9UVhWWHmjAsopZB/OOeciUO5JSFINSYfmTAPdgYXAVCCnhdsg4JVweiowMGwl1wX4MaxS\nmw50l1QnbOXWHZgeztsoqUvYKm5gvm0VtA/nnHMRiKI67gjgpbDVdBXgWTN7U9LHwL8kXQosA84N\nl3+doGXcUmArcDGAmf0g6U7g43C5O8zsh3D6KuAp4GDgjfABMDLOPpxzzkWg3JOQmX0NtCugfB1w\negHlBlwdZ1vjgfEFlGcCbRPdh3POuWgkUxNt55xzFYwnIeecc5HxJOSccy4ynoScc85FxpOQc865\nyHgScs45FxlPQs455yLjScg551xkPAk555yLjCch55xzkfEk5JxzLjKehJxzzkXGk5BzzrnIeBJy\nzjkXGU9CzjnnIuNJyDnnXGQ8CTnnnIuMJyHnnHOR8STknHMuMp6EnHPORcaTkHPOuch4EnLOORcZ\nT0LOOeci40nIOedcZDwJOeeci4wnIeecc5HxJOSccy4ynoScc85FxpOQc865yHgScs45FxlPQs4l\noT179tChQwf69OkDwNixYzn66KORxNq1awtd99tvv6VmzZr89a9/BWD79u0cd9xxtGvXjjZt2nDr\nrbfmLnvBBRdw7LHH0rZtWy655BJ27dpVdi/KuQJ4EnIuCf3tb3+jVatWuc+7du3KW2+9RZMmTYpc\n97rrrqNXr165zw866CBmzZrFZ599xrx583jzzTf58MMPgSAJLV68mAULFrBt2zaeeOKJ0n8xzhXC\nk5BzSSY7O5tp06Zx2WWX5ZZ16NCBpk2bFrnuyy+/TLNmzWjTpk1umSRq1qwJwK5du9i1axeSAPjN\nb36DJCRx3HHHkZ2dXbovxrkieBJyB6T81Vmxnn/+edq0aUOlSpXIzMzMLV+3bh2nnnoqNWvWZMiQ\nIXnW2blzJ4MHD+aYY46hZcuWvPjii2UW+7Bhw7jvvvuoVKl4H8/Nmzdz77335qluy7Fnzx7at29P\nWloa3bp14/jjj88zf9euXUycOJGePXvuV+wuORR2/P/www9069aNFi1a0K1bN9avXw/A+vXr6dev\nHxkZGRx33HEsXLgQgC+//JL27dvnPmrVqsXo0aNLLVZPQu6AlL86K1bbtm2ZMmUKJ510Up7y6tWr\nc+edd+ZeS4l11113kZaWxv/+9z8WLVrEySefXCZxv/baa6SlpdGpU6dir3vbbbdx7bXX5p71xKpc\nuTLz5s0jOzubOXPm5H7B5Ljqqqs46aST+NWvflXi2F3yKOz4HzlyJKeffjpLlizh9NNPZ+TIkQDc\nfffdtG/fnvnz5/P0008zdOhQAI499ljmzZvHvHnzmDt3Locccgj9+vUrtVg9CbkDTkHVWbFatWrF\nscceu095jRo1OPHEE6levfo+88aPH8/NN98MQKVKlTj88MNLN+jQ+++/z9SpU2natCnnn38+s2bN\n4sILL0xo3Y8++ojhw4fTtGlTRo8ezd13383YsWPzLFO7dm1OPfVU3nzzzdyy22+/nTVr1jBq1KhS\nfS0uGkUd/6+88gqDBg0CYNCgQbz88ssALFq0iNNOOw2Ali1bkpWVxapVq/KsO3PmTJo3b57QtclE\neRJyB5ySVmfFs2HDBgD+9Kc/0bFjR84555x9Ppyl5Z577iE7O5usrCwmT57MaaedxqRJk+IuP2fO\nHAYOHAjAu+++S1ZWFllZWQwbNoxbbrmFIUOGsGbNmtzXsG3bNmbMmEHLli0BeOKJJ5g+fTrPPfdc\nqb1fLlpFHf+rVq2ifv36APz85z/PPZbbtWvHlClTgOC4WrZs2T7XCCdPnsyAAQNKNd4qpbo15yIW\nW5319ttvl8o2d+/eTXZ2NieccAKjRo1i1KhR3HDDDUycOLFUth+r6YhpudPbv53PxsWraTpiGhsz\np7LxoxfZs2U9P292LAcf1ZnDel3DlsXvsT1rTZ71ADa89z9U9WDGrp3GztXfsHbag2B7OSatBuee\ne27utYIrrriCJk2a8Mtf/hKA/v378+c//7nUX5crH8U9/nMapQCMGDGCoUOH0r59e9LT0+nQoQOV\nK1fOXXbnzp1MnTqVe+65p1RjrpBJSFJP4G9AZeAJMxsZcUiulORUZ73++uts376djRs3cuGFFxZ6\nNlGUww47jEMOOYT+/fsDcM455/Dkk0+WVshxVW+cQfXGGQDU6nwGtTqfsc8yO1f+j0M79t6nvPaJ\nF+ROV0trRoOLxwCwcGTeZXfv3l2aIbuIJXL8H3HEEaxcuZL69euzcuVK0tLSAKhVqxb/+Mc/ADAz\nmjVrxlFHHZW73htvvEHHjh054ogjSjXmCnf+LakyMA7oBbQGBkhqHW1UrrTEq866+eabeemll0q0\nTUn07ds395flzJkzad06OQ6ZOqdeQrW0ZlGH4ZJEIsf/GWecwYQJEwCYMGECZ555JhBUO+/cuRMI\nqmlPOukkatWqlbvt5557rtSr4qBingkdByw1s68BJE0GzgQWRRqVKxXxqrNWvzCTn2UfxrUfTWPr\n/z7ghxmPsmfbjxx/cjeqpTXjiPPuBCD74UuwnVuxPbt5ZMJk0s67k2qHN2Z3nR68POgajqlTiXr1\n6uX+YnQuWeSvko13/O/Z2ZG1j4zkT/eNoUqtNA4/cwT/GjGNHSu+CKptJaoe3pjDeg3N3ebendvZ\nMmMGjz76aKnHLTMr9Y0mM0lnAz3N7LLw+UXA8WY2JGaZwcDg8OmxwJdlGNLhQOH9sCS3VIm/BbCk\ngPJUiT+eVI4/lWOH1Iq/oOO/rONvYmb1ilqoIp4JFcnMHgMeK499Sco0s87lsa+y4PFHK5XjT+XY\nweMvLRXumhCwAjgy5nmjsMw551w5q4hJ6GOghaRmkqoB5wNTI47JOecqpApXHWdmuyUNAaYTNNEe\nb2afRxhSuVT7lSGPP1qpHH8qxw4ef6mocA0TnHPOJY+KWB3nnHMuSXgScs45FxlPQs455yJT4Rom\nJANJlYB2QANgG7DQzFZHG1XxSKoBbDezPVHHUlyS6vDTe59lZnsjDqlYUjV+SdWBPsCviDn2gWkR\nNw5KiKRGBK1p94kfeCPZ/w+SfglcSBB/ffLGP8nMfowkLm+YUH4kNQduAn5NcPfyGqA6cAywFXgU\nmJCMB3OYOM8HLgB+AewADiK443oa8KiZLY0uwsJJ+hlwNTAAqMZP7/0RwIfAQ2Y2O7oIC3cAxH87\nQQJ6G5gLrOanY//UcPp6M5sfVYyFkfQPoCHwGpDJvvF3AkaY2X8iC7IQkt4AvgNeoeD4+wKjzKzc\nb1fxJFSOJD0HPAy8a/neeElpwO+A9WY2IYr4CiPpHeAtgoN4YU6ilFSX4CD+HfCSmZW8u+oyJGkG\n8DTwqpltyDevE3ARsMDMyr577BI4AOLvbWbTCpmfBjQ2s8x4y0RJUlszW1jI/GoE8SflDzFJh5tZ\noV30JLJMWfAk5BIiqaqZ7drfZZxLdeEPL8zsh6hjORB4EipnkloS9NrdMCxaAUw1sy+iiyoxCka/\nOo68sc/Jf1aXaiS1NLPFUcdRlLBKrid53//p+c+MklEY+83Ab4E0wAiqhF4BRib7a5DUGLgPOB3Y\nAAioBcwiqIbLii66/SNpgZmlR7b/FP/+SCmSbiKo058M5Iybm3Oxc3IyD64nqTvwEMG1rJy+9hoB\nRwNXmdm/o4ptf0n61swaRx1HYSQNBG4F/k3e978bcLuZPR1VbImQNJ3gC3uCmX0flv0cGAScbmbd\no4yvKJL+C4wGXshpjBOOTXYOMMzMukQZX1Ek9Y83C3gkkd6uy4onoXIk6X9Am/xVVmF98udm1iKa\nyIom6QugV/5ffJKaAa+bWatIAkuQpDHxZgGDzKxWnPlJQdKXBEOO5L8eVAf4yMyOiSayxEj60syO\nLe68ZCFpSbzPZ2HzkoWkXcAzBGeg+Z1tZoeWc0i5vIl2+dpL0LRzWb7y+uG8ZFaFn87eYq0AqpZz\nLCVxMXA9Qau+/Ep/uMjSJwr+Atkbzkt2yyQNJzgTWgUg6Qjg98DyKANL0FxJDwET+CneIwnO5D6N\nLKrEzQf+WlDjCkm/jiCeXJ6EytcwYKakJfx0IDcmqNIaEnet5DAe+DgciTb2Q3g+kJQtsvL5mKBV\n3wf5Z0i6rfzDKba7gE8k/Zu8x0434M7IokrcecAI4J2wJRzAKoIe7M+NLKrEDQQuBW7np2ty2cCr\npMbxPwzYGGdev/IMJD+vjitn4f02+S/uf5wKN31KakXBjSqSfmj0sEXTdjPbGnUsJRVWvfVg34YJ\n66OLyrn940nIOedcZLzvOOecc5HxJOSccy4ynoScc5GR1FlSg6jjKClJZ0o6Puo4SioZ4vfWcUlA\n0lvALmCcmb0WdTzFIelu4EfgCTNbF3U8xXUAxD+BoPPbcYX1bZbE/ghkSPqfmZ0XdTAlcDyQLqmK\nmfWKOpgSiDx+b5iQBMJfgvWBLmY2Lup4ikPSb4HmQDszGxh1PMV1AMT/C4Km2seZ2U1RxxOPpMZm\n9m0h8w81s03lGZNLDp6EXEIkDTGzsVHHUVIHQPx3m9ktUcdRUpI+MbOOUcdR2iR1A4abWbeoYymJ\nZIjfrwmVI0mzJc2K85gZdXxFuCTqAPZTqsffM+oA9lMq9OoQl6TTJP1P0mZJkySlS8oERhIMz5LU\nkjl+vyZUvm4ooKwLMJygR2Hn4qkc3qxa4Jd5Cgwr0LCQ/vsws2vKM5gSeAAYDPwX6BX+HZFCZ9dJ\nG79Xx0VE0snAnwhGN7zLzN6IOKRCSdpNcAF8n1mApUAHoKke/w6CHhIKSkJmZkeVc0jFImkZ8Od4\n85NxIMdY+asTU6HT1VjJHL+fCZUzST2A/0fQkeZdyTwkcz4LzKxD1EHsh1SPf1GKx78u2RNNEWrn\nGw6hSuxzM5sSQUzFkbTxexIqR5I+BuoB9xOcDiMp99eJmX0SUWjOlbWdUQewn/4D9I3z3IBkT0JJ\nG78nofK1BdgMnB0+YhlwWrlHlLjnow5gP6V6/H+LOoD9dG44OmmBCmu+nQzM7PdRx7A/kjl+vybk\nEiLp7+Qdz8aAtcBsM3svmqgSdwDE/yoFjycEgJmdUY7hFJukBQTxx17TMoKagTQzqxxJYAmSNNrM\nhoXTQ83sbzHznkrmL3lI7vj9TKgcFTLELpD09cqZBZTVBe6X9E8zG13eARVTqsf/16gD2B9mlh77\nXFJT4Cbg18DdEYRUXCfFTA8i75lpRjnHUhJJG78nofLVt5B5SV2vHO+isqRHgA+ApP4SPwDifyfq\nGEqDpBbA/xF0F/MAcE3+4e6TlOJMp4qkjd+TUDkys4sTWU7SoFRpSWRm26SkOqaLJVXiD7+8bwHW\nA6OAx4FfAV8Bl5pZQWd6SUNSW4Lk0wa4jyDmpB/IMUal8D6tSjHTOQdOUlclhpI2fr8mlIRSpYsT\nSVWAi4D+ZlbYWV5SSqX4Jb0HPA3UAq4lGK75VYJE9BczS+qenCXtIRiWfBqwT/JJ9ptVJWUBe0nd\n+7SySNL4PQklIUmfJts9IZI2se+F5a3AO8AwM/suksASdADEP8/M2ofTS83s6ILmJStJv6fwhhUp\ncebvSp9XxyWnpPtlYGaHRh3D/kj1+Al+xebYWMi8pGRmT8WbF56RJrXYDnAltTGzz6OOqTiSOX4/\nE0pCyXgmBCCpGnABQb0+wOfAs2a2I7qoEpfK8UvaCiwlOJNrHk4TPj/KzGpEFVsiJL1nZieG0xPN\n7KKYeUlf/RwbYyrEm18yx5/0v0AqqPejDiA/Sa2BqQSxzQ2LTwH+T9IZZrYoqtgSkerxA62iDmA/\nxSbJNvnmJX/LkLxSLd78kip+T0LlSNJ1hc03s1Hh3yHlE1Gx/B240sxmxBZK+jUwDjg1kqgSl+rx\nH2xmiwEkHRR79iapC7AsssgSU1iVSypUx9SW1I+gdVmt/Pf8Jfk9fpDE8Xt1XDmSdGvM08uBR2Pn\nm9nt5RtR4iQtNrOWceZ9YWZJ/Uv9AIg/bnVKslWvFETS18D1BF+C9/PTsCYC7jOz5lHFlghJ/yhk\ntplZUo9Xlczx+5lQOYpNMpJ+m8xJpwCV8v8CB5BUndQ4jlI9/sJuNkyq6pU43gHOiJnO35lmUkv0\nHr9klczxp8KH70CVaqegTwMvSrrazJZBbtcrY4CJEcaVqFSPP3+/d/HmJaVk/hJMhKS+wPyYY+fP\nwFkE1aBDzeybKOMrSjLH79VxEUmFKpT8JA0hGAX2kLBoC/BXM/t7dFElLpXjl7QamExw1nNeOE34\n/FwzOyKq2BKR6PXQZCVpPtDFzLZK6kPQa8UAoANwjpn1iDTAIiRz/J6EylFMT8IAR5O3ma2ZWSp0\nhIikQwHMbFMB85K+y6FUjF/SoMLmJ2PMsSTtBeYBbxAM6JinCjHZq6YlfWZm7cLp8cCXZnZv+Dzp\nf1Amc/xeHVe++kQdQGko6Ms7xlAgqb8QUzH+ZE8yCehA8Mu7N0ET+eeAmZY6v4IlqSZBLxunAw/F\nzKseTUjFkrTxV4py5xWNmS2LfRAMcNcRODynrvYAkAoXyQuTtPFLGiTpE0lbwkempIFRx5UIM/vM\nzEaE3Qs9CZwJLJKU1OMgxRhNcCaXCXyR02GspA7AyigDS1DSxu9nQuVI0mvACDNbKKk+8AnBQdFc\n0mOW/GPaJCJVftnGk5Txh9Vxw4DrCI4bEfyAuV+SmVkqNK5AUj2Cs6J0IBtYHW1EiTGz8ZKmA2nA\nZzGzvgeSvtFFMsfv14TKkaTPzaxNOH0L0NLMBobXKN5PlWtChUnWLocSlazxS/oQON/MsvKVNwUm\nm1mXCMJKmKRLgHMJqn5eAP5lZimRgABUyNDkkPzDkydz/H4mVL5iB+86nWBMGMxsU3jh9kCQdF0O\nFVOyxl8rfwICMLMsSbUiiKe4ngAWEjQJ7gF0V8w4Tpbkw5MTDEERd3hykn9MoaSN35NQ+Vou6Y8E\n1RAdgTcBJB0MVI0ysKKkeJdDKR8/sK2E85JFsneLVChL8eHJkzl+T0Ll61LgDoJ//HlmtiEs7wIU\n1q1GMogdCmGfLodSQKrH3yq81yM/AUk9oFqoC8E9Wak0muo+lLrDkwPJGb9fE0oCYdcxfc3s+ahj\nSUSyXjdJVCrGL6lJYfOTvXWlpLHAicDVZpasVZ5xad/hyZ9LpYSazPF7EoqIpMoEdeMDgO7Au2Z2\ndrRRJSbqm9v2VyrGr7AJ3P4uEyVJHYGxwBfAw8QMxmdmn0QVVyKU+sOTJ238Xh1XziSdDPwO+A0w\nB+gKNDOzrZEG5pLdbEkvAq/EtmRSMFDficAgYDbwVDThFc3MPglbhb5IMDBfTsI04LTIAktMUveS\nnYCkjd/PhMqRpGzgW4JfgS+HreK+MbNmEYdWpFTvcugAiL86wRfJBUAzYANBc+fKwL+Bh8zs0+gi\nLJykNIJrEEcBV5nZZ0WskhJSrSo9v2SI35NQOZI0GvgtQVPVZ4FXgAVmlvQXlg+AaxIpHX8sSVWB\nw4FtMY1bkpqkb4B7gMeTucowEalclQ7JF78noXKm4OaIUwgOgN8APyNoNfe6mW2OMLRikXQYcBLw\nrZnNLWr5ZJPq8acaSfXMbE0B5UcS3IR7fwRhFUucqvSjUqUqPVnj977jypkFZpvZYIJqlQEE/Whl\nRRpYESS9FrawIexyaCFB9dBEScMiDS4BqR5/qotNQJLqSbpK0rvA20BSD0MBuVXp9wDvAa3N7CyC\nM9FUSUBJG78noXIUfvha5zw3s11m9hrBzWLJ3lqrmZktDKcvBmaYWV+C+w2S9qJnjFSPP6VJOjTs\ngHU6wa/w5gT/k+ZmdkMRqyeDF4AGBGM59ZVUgyTtZzCOpI3fk1D5+jtBXX5+dYGR5RxLceXvcuh1\nyB0WIRW6HEr1+FPdaoJk/xeCKqDrgZ3RhpQ4MxtGUHPxAEF1+pdAPUnnhkMkJLVkjt+vCZUjSZlm\n1jnOvIVm1ra8Y0qUpFcJWmFlA+MJfsVuCLscyszpmDVZpXr8qS6s8jwfqEEwltA/Cc5Gk75RTkHC\nxiE5F/d7mFlBPy6TVjLF72dC5evQQuYldd9xBI0n2gC/J/W6HILUjz+lmdnosKfvM8Oil4EGkm6S\ndEyEoZVITlW6mV0A9I86nuJKpvj9TKgcSZoGjDOz1/OV9yLow6lXNJGVXDLcZ7A/Uj3+VBY2FBlA\n8KPg6KjjKUzYrPlcoCHwpgVjgvUBbgEOTvZuoJI5fk9C5Sj8xfca8AHBEMcAnYFfAn3M7H9RxVYc\nyXafQXGlevypStJvCW4UXmBm06OOpzgkPQUcSdCo4njgO4LP7ggzeznC0BKSzPF7Eipnkg4iaKuf\nc/3nc+BZs//f3t2E2FXecRz//pqFiS2IGqUkGzEqURMT1AouDNZFFBHciIJvSGhFuyhEFFoqNsWF\nBWmJV44AAAXCSURBVC2ilhHfMKAL8RUVF75hfKE2BN+SoVlUYkSwUqIWDTEVk18X57nJzTiTOXdI\n5pxn+vvAhZlzzuT8bhb3f5/nnPN/vLu7VO309TmDtmrPXzNJYzTToX+juTHkRdu3d5uqPUnjwBm2\n95bR8xfAEttfdhytlT7nT++4WVQaTP6Xg1yD6GsTygkth24eajlUxQd47fnngFXACtt7JB0JvA1U\nU4SA723vBbC9W9K2PnyAj6C3+VOEZlfNTSifpmk5dAWwR9Lz9OQ5g5Zqz1+7712WDrC9S8PLqtZh\n6dB6TgKWlN+r6D1Ij/NnOm4WTdGEcgHNXYo1NKGsuuVQ7flrJmkXBzaNXVJ+7/xDsI3aew/2OX+K\nUEdqbEI5rE/PGcxE7flr0+cPwTYkvWJ7ddc5ZqrP+VOEohVJxwHH2f7HhO2nA98OTy/2Ue35a9fm\nWmdfr4dCnavxDutz/lwTirbuA8Ym2X4MzbLBV85unJHVnr92NV8PBThK0pQPddp+djbDzEBv82ck\nFK3U3HII6s9fuymuh9a0KN+XNOt/TXZDhW33uglun/NnJBRt1dxyCOrPX7XyHNwYMFbp9dBP+15o\nptHb/OkdF219LOniiRtLy6FtHeQZVe3554zSt+xfFRUgmHwEUZPe5s90XLRSe8uh2vNHtyQtt71l\nmmP6fGNFb/OnCEVrNbccgvrzR3ckbQCmvbHC9vpOAk6jz/lThKKVOXCLbdX5o1tz4MaK3uZPEYpW\n+vxNqo3a80d/VHpjxT59y58iFK3MgZZDVeePmKtShGJkffsmNara80fMJSlCERHRmTwnFBERnUkR\nioiIzqQIRYxA0gJJb0qaJ+kESbPa+FTSw5JOO8j+9ZI+kfRhef32MOXYIOns8vNrko4+HOeJuS9F\nKGI0a4BnyyqhJzDL3bdt/2richSTuMX2yvK6d+JOSfMOcazHgN8c4n8z/k+kCEWM5iqabsQAfwbO\nKyOOtZLekrRycKCkdyStkLRO0mOS3pX0T0m/HjrmFkmbJG2W9KfpTj4YgZSR2HpJ45K2SFo7zd/t\nlPQXSR8B50q6rZx3XNKDg+W2J4xwFkraXn5eIOkJSVslPUdze/vACzSLA0aMLEUooqXyYOuJtreX\nTb8D3i4jjruBR4DryrGnAPNtf1SOPQO4gKZX3W2SFklaDZwMnAOsBM6StKplnJXAYtvLbC8HHh3a\nd+fQdNzysu2nwEbbK2y/A/zV9i/KEhYLgEumOd+NwC7bpwJ/BM4a7LD9NXCEpGNbZo/YJ0Uoor2F\nNA+5TuUp4JLyHNIaDlyg7Xnb39neQbN42znA6vL6AHgfWEpTlNrYBpwo6T5JFwHfDO0bno4bNK3c\nQ9MxYuCXkjZK2kJTHE+f5nyrgMcBbG8GNk/Y/29gUcvsEftkPaGI9r6j6bc1Kdu7JL0KXApcztBo\nAZj4QJ5p2uvfYfuBUYPY/lrSCuBC4IZyvoOtF7O7XMcadI8YA862/Zmkdex/Xz+w/8vplO91EvNp\n/n8iRpKRUERLZdppXvkQB/iWHy+W9zBwL7CpHD9wqaT5ZcrqfGAT8DKwRtLPACQtlnR8+fl1SYun\nyiJpIfAT288AtwJnjvBWBvl3lHNfNrRvO/uL5/D2tyg3YUhaRjO9OMgi4OflbyNGkpFQxGheoWl4\n+hrNlNSecrF/ve27bb8n6RsOvEZDOfYNmim9221/Dnwu6VTg3XJfwE7gakk7gJOArw6SYzHwqKTB\nF8nft30Dtv8j6SFgHPiCpiAO3AU8Kel64KWh7feX820FtrJ/TSZoitbfbf/QNkPEQNr2RIxA0pnA\nWtvXTLF/EbABWGp7b9m2Dthp+66W51gGrLF90yEJfZhJugd4wfbrXWeJ+mQ6LmIEtt8H3pjsWRtJ\n1wIbgT8MCtAMzzFeSwEqxlOAYqYyEoqIiM5kJBQREZ1JEYqIiM6kCEVERGdShCIiojMpQhER0Zn/\nAVkX27zx072vAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fec5b3efdd8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "# 查看转账类型和欺诈标记的记录\n",
    "ax = raw_data.groupby(['type', 'isFraud']).size().plot(kind='bar')\n",
    "ax.set_title('# of transactions vs (type + isFraud)')\n",
    "ax.set_xlabel('(type, isFraud)')\n",
    "ax.set_ylabel('# of transaction')\n",
    "\n",
    "# 添加标注\n",
    "for p in ax.patches:\n",
    "    print(p)\n",
    "    ax.annotate(str(format(int(p.get_height()), ',d')), (p.get_x(), p.get_height()*1.01))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAmcAAAJXCAYAAADWyPO8AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fec5b805438>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axs = plt.subplots(2, 2, figsize=(10, 10))\n",
    "transfer_data = raw_data[raw_data['type'] == 'TRANSFER']\n",
    "\n",
    "a = sns.boxplot(x='isFlaggedFraud', y='amount', data=transfer_data, ax=axs[0][0])\n",
    "axs[0][0].set_yscale('log')\n",
    "\n",
    "b = sns.boxplot(x='isFlaggedFraud', y='oldbalanceDest', data=transfer_data, ax=axs[0][1])\n",
    "axs[0][1].set(ylim=(0, 0.5e8))\n",
    "\n",
    "c = sns.boxplot(x='isFlaggedFraud', y='oldbalanceOrg', data=transfer_data, ax=axs[1][0])\n",
    "axs[1][0].set(ylim=(0, 3e7))\n",
    "\n",
    "d = sns.regplot(x='oldbalanceOrg', y='amount', data=transfer_data[transfer_data['isFlaggedFraud'] ==1], ax=axs[1][1])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/airflow/jupyter/lib/python3.4/site-packages/ipykernel_launcher.py:3: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/indexing.html#indexing-view-versus-copy\n",
      "  This is separate from the ipykernel package so we can avoid doing imports until\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "       type     amount  oldbalanceOrg  newbalanceOrig  oldbalanceDest  \\\n",
      "0  TRANSFER     181.00          181.0             0.0             0.0   \n",
      "1  CASH_OUT     181.00          181.0             0.0         21182.0   \n",
      "2  CASH_OUT  229133.94        15325.0             0.0          5083.0   \n",
      "3  TRANSFER  215310.30          705.0             0.0         22425.0   \n",
      "4  TRANSFER  311685.89        10835.0             0.0          6267.0   \n",
      "\n",
      "   newbalanceDest  isFraud  \n",
      "0            0.00        1  \n",
      "1            0.00        1  \n",
      "2        51513.44        0  \n",
      "3            0.00        0  \n",
      "4      2719172.89        0  \n",
      "        type     amount  oldbalanceOrg  newbalanceOrig  oldbalanceDest  \\\n",
      "0   TRANSFER     181.00         181.00            0.00            0.00   \n",
      "1   CASH_OUT     181.00         181.00            0.00        21182.00   \n",
      "2   CASH_OUT  229133.94       15325.00            0.00         5083.00   \n",
      "3   TRANSFER  215310.30         705.00            0.00        22425.00   \n",
      "4   TRANSFER  311685.89       10835.00            0.00         6267.00   \n",
      "5   CASH_OUT  110414.71       26845.41            0.00       288800.00   \n",
      "6   CASH_OUT   56953.90        1942.02            0.00        70253.00   \n",
      "7   CASH_OUT    5346.89           0.00            0.00       652637.00   \n",
      "8   CASH_OUT   23261.30       20411.53            0.00        25742.00   \n",
      "9   TRANSFER   62610.80       79114.00        16503.20          517.00   \n",
      "10  CASH_OUT   82940.31        3017.87            0.00       132372.00   \n",
      "11  CASH_OUT   47458.86      209534.84       162075.98        52120.00   \n",
      "12  CASH_OUT  136872.92      162075.98        25203.05       217806.00   \n",
      "13  CASH_OUT   94253.33       25203.05            0.00        99773.00   \n",
      "14  TRANSFER   42712.39       10363.39            0.00        57901.66   \n",
      "15  TRANSFER   77957.68           0.00            0.00        94900.00   \n",
      "16  TRANSFER   17231.46           0.00            0.00        24672.00   \n",
      "17  TRANSFER   78766.03           0.00            0.00       103772.00   \n",
      "18  TRANSFER  224606.64           0.00            0.00       354678.92   \n",
      "19  TRANSFER  125872.53           0.00            0.00       348512.00   \n",
      "\n",
      "    newbalanceDest  isFraud  typeCategory  \n",
      "0             0.00        1             1  \n",
      "1             0.00        1             0  \n",
      "2         51513.44        0             0  \n",
      "3             0.00        0             1  \n",
      "4       2719172.89        0             1  \n",
      "5          2415.16        0             0  \n",
      "6         64106.18        0             0  \n",
      "7       6453430.91        0             0  \n",
      "8             0.00        0             0  \n",
      "9          8383.29        0             1  \n",
      "10        49864.36        0             0  \n",
      "11            0.00        0             0  \n",
      "12            0.00        0             0  \n",
      "13       965870.05        0             0  \n",
      "14        24044.18        0             1  \n",
      "15        22233.65        0             1  \n",
      "16            0.00        0             1  \n",
      "17       277515.05        0             1  \n",
      "18            0.00        0             1  \n",
      "19      3420103.09        0             1  \n"
     ]
    }
   ],
   "source": [
    "used_data = raw_data[(raw_data['type'] == 'TRANSFER') | (raw_data['type'] == 'CASH_OUT')]\n",
    "# 丢掉不用的数据列\n",
    "used_data.drop(['step', 'nameOrig', 'nameDest', 'isFlaggedFraud'], axis=1, inplace=True)\n",
    "# 重新设置索引\n",
    "used_data = used_data.reset_index(drop=True)\n",
    "print(used_data.head())\n",
    "\n",
    "# 将type转换成类别数据，即0, 1\n",
    "type_label_encoder = preprocessing.LabelEncoder()\n",
    "type_category = type_label_encoder.fit_transform(used_data['type'].values)\n",
    "used_data['typeCategory'] = type_category\n",
    "\n",
    "print(used_data.head(20))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7fec233de400>"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAaYAAAFFCAYAAACnoUkvAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nE5OZmRXl/wGTzUCGtru/YAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fec599bf898>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "# 查看变量间的相关性\n",
    "sns.heatmap(used_data.corr())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZMAAAGhCAYAAACkrlnHAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nMJEkFfv/1j3g0nUWcrkAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fec5b36b780>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZMAAAEFCAYAAAAmIwo/AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nD7xn327/RG8PAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fec5b329898>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# 查看转账类型记录个数\n",
    "ax = used_data['type'].value_counts().plot(kind='bar', title=\"Transaction Type\", figsize=(6,6))\n",
    "for p in ax.patches:\n",
    "    ax.annotate(str(format(int(p.get_height()), ',d')), (p.get_x(), p.get_height()*1.01))\n",
    "plt.show()\n",
    "\n",
    "# 查看转账类型中欺诈记录个数\n",
    "ax = pd.value_counts(used_data['isFraud'], sort = True).sort_index().plot(kind='bar', title=\"Fraud Transaction Count\")\n",
    "for p in ax.patches:\n",
    "    ax.annotate(str(format(int(p.get_height()), ',d')), (p.get_x(), p.get_height()))\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "      amount  oldbalanceOrg  newbalanceOrig  oldbalanceDest  newbalanceDest  \\\n",
      "0     181.00          181.0             0.0             0.0            0.00   \n",
      "1     181.00          181.0             0.0         21182.0            0.00   \n",
      "2  229133.94        15325.0             0.0          5083.0        51513.44   \n",
      "3  215310.30          705.0             0.0         22425.0            0.00   \n",
      "4  311685.89        10835.0             0.0          6267.0      2719172.89   \n",
      "\n",
      "   typeCategory  \n",
      "0             1  \n",
      "1             0  \n",
      "2             0  \n",
      "3             1  \n",
      "4             1  \n",
      "0    1\n",
      "1    1\n",
      "2    0\n",
      "3    0\n",
      "4    0\n",
      "Name: isFraud, dtype: int64\n"
     ]
    }
   ],
   "source": [
    "# 准备数据\n",
    "feature_names = ['amount', 'oldbalanceOrg', 'newbalanceOrig', 'oldbalanceDest', 'newbalanceDest', 'typeCategory']\n",
    "X = used_data[feature_names]\n",
    "y = used_data['isFraud']\n",
    "print(X.head())\n",
    "print(y.head())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[  1.81000000e+02   1.81000000e+02   0.00000000e+00   0.00000000e+00\n",
      "    0.00000000e+00   1.00000000e+00]\n",
      " [  1.81000000e+02   1.81000000e+02   0.00000000e+00   2.11820000e+04\n",
      "    0.00000000e+00   0.00000000e+00]\n",
      " [  2.80600000e+03   2.80600000e+03   0.00000000e+00   0.00000000e+00\n",
      "    0.00000000e+00   1.00000000e+00]\n",
      " ..., \n",
      " [  5.56347680e+05   0.00000000e+00   0.00000000e+00   2.35494738e+06\n",
      "    2.91129506e+06   0.00000000e+00]\n",
      " [  1.40601560e+05   0.00000000e+00   0.00000000e+00   1.62351880e+06\n",
      "    1.76412036e+06   0.00000000e+00]\n",
      " [  1.43448170e+05   0.00000000e+00   0.00000000e+00   8.64368630e+05\n",
      "    1.00781681e+06   0.00000000e+00]]\n",
      "非欺诈记录比例:  0.5\n",
      "欺诈记录比例:  0.5\n",
      "欠采样记录数:  16426\n"
     ]
    }
   ],
   "source": [
    "# 处理不平衡数据\n",
    "# 欺诈记录的条数\n",
    "number_records_fraud = len(used_data[used_data['isFraud'] == 1])\n",
    "# 欺诈记录的索引\n",
    "fraud_indices = used_data[used_data['isFraud'] == 1].index.values\n",
    "\n",
    "# 得到非欺诈记录的索引\n",
    "nonfraud_indices = used_data[used_data['isFraud'] == 0].index\n",
    "\n",
    "# 随机选取相同数量的非欺诈记录\n",
    "random_nonfraud_indices = np.random.choice(nonfraud_indices, number_records_fraud, replace=False)\n",
    "random_nonfraud_indices = np.array(random_nonfraud_indices)\n",
    "\n",
    "# 整合两类样本的索引\n",
    "under_sample_indices = np.concatenate([fraud_indices, random_nonfraud_indices])\n",
    "under_sample_data = used_data.iloc[under_sample_indices, :]\n",
    "                          \n",
    "                          \n",
    "X_undersample = under_sample_data[feature_names].values\n",
    "y_undersample = under_sample_data['isFraud'].values\n",
    "\n",
    "# 显示样本比例\n",
    "print(\"非欺诈记录比例: \", len(under_sample_data[under_sample_data['isFraud'] == 0]) / len(under_sample_data))\n",
    "print(\"欺诈记录比例: \", len(under_sample_data[under_sample_data['isFraud'] == 1]) / len(under_sample_data))\n",
    "print(\"欠采样记录数: \", len(under_sample_data))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[ -1.84334244  -1.32551771   0.16288746 ..., -19.78446999   1.02915705\n",
      "  -7.47494863]\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYoAAAEWCAYAAAB42tAoAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n2ZZ2AAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fec234615c0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# 选用逻辑回归模型进行预测\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.metrics import roc_curve, auc\n",
    "\n",
    "# 分割训练集和测试集\n",
    "X_train, X_test, y_train, y_test = train_test_split(X_undersample, y_undersample, test_size=0.3, random_state=0)\n",
    "\n",
    "lr_model = LogisticRegression()\n",
    "lr_model.fit(X_train, y_train)\n",
    "\n",
    "y_pred_score = lr_model.decision_function(X_test)\n",
    "print(y_pred_score)\n",
    "\n",
    "fpr, tpr, thresholds = roc_curve(y_test, y_pred_score)\n",
    "roc_auc = auc(fpr,tpr)\n",
    "\n",
    "# Plot ROC\n",
    "plt.title('Receiver Operating Characteristic')\n",
    "plt.plot(fpr, tpr, 'b',label='AUC = %0.2f'% roc_auc)\n",
    "plt.legend(loc='lower right')\n",
    "plt.plot([0,1],[0,1],'r--')\n",
    "plt.xlim([-0.1,1.0])\n",
    "plt.ylim([-0.1,1.01])\n",
    "plt.ylabel('True Positive Rate')\n",
    "plt.xlabel('False Positive Rate')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "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": "3.4.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}