Pro.Fraud_Detection.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\IPython\\core\\interactiveshell.py:3058: DtypeWarning: Columns (0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30) have mixed types. Specify dtype option on import or set low_memory=False.\n",
      "  interactivity=interactivity, compiler=compiler, result=result)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</th>\n",
       "      <th>...</th>\n",
       "      <th>21</th>\n",
       "      <th>22</th>\n",
       "      <th>23</th>\n",
       "      <th>24</th>\n",
       "      <th>25</th>\n",
       "      <th>26</th>\n",
       "      <th>27</th>\n",
       "      <th>28</th>\n",
       "      <th>29</th>\n",
       "      <th>30</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>Time</td>\n",
       "      <td>V1</td>\n",
       "      <td>V2</td>\n",
       "      <td>V3</td>\n",
       "      <td>V4</td>\n",
       "      <td>V5</td>\n",
       "      <td>V6</td>\n",
       "      <td>V7</td>\n",
       "      <td>V8</td>\n",
       "      <td>V9</td>\n",
       "      <td>...</td>\n",
       "      <td>V21</td>\n",
       "      <td>V22</td>\n",
       "      <td>V23</td>\n",
       "      <td>V24</td>\n",
       "      <td>V25</td>\n",
       "      <td>V26</td>\n",
       "      <td>V27</td>\n",
       "      <td>V28</td>\n",
       "      <td>Amount</td>\n",
       "      <td>Class</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>-1.3598071336738</td>\n",
       "      <td>-0.0727811733098497</td>\n",
       "      <td>2.53634673796914</td>\n",
       "      <td>1.37815522427443</td>\n",
       "      <td>-0.338320769942518</td>\n",
       "      <td>0.462387777762292</td>\n",
       "      <td>0.239598554061257</td>\n",
       "      <td>0.0986979012610507</td>\n",
       "      <td>0.363786969611213</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.018306777944153</td>\n",
       "      <td>0.277837575558899</td>\n",
       "      <td>-0.110473910188767</td>\n",
       "      <td>0.0669280749146731</td>\n",
       "      <td>0.128539358273528</td>\n",
       "      <td>-0.189114843888824</td>\n",
       "      <td>0.133558376740387</td>\n",
       "      <td>-0.0210530534538215</td>\n",
       "      <td>149.62</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>1.19185711131486</td>\n",
       "      <td>0.26615071205963</td>\n",
       "      <td>0.16648011335321</td>\n",
       "      <td>0.448154078460911</td>\n",
       "      <td>0.0600176492822243</td>\n",
       "      <td>-0.0823608088155687</td>\n",
       "      <td>-0.0788029833323113</td>\n",
       "      <td>0.0851016549148104</td>\n",
       "      <td>-0.255425128109186</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.225775248033138</td>\n",
       "      <td>-0.638671952771851</td>\n",
       "      <td>0.101288021253234</td>\n",
       "      <td>-0.339846475529127</td>\n",
       "      <td>0.167170404418143</td>\n",
       "      <td>0.125894532368176</td>\n",
       "      <td>-0.00898309914322813</td>\n",
       "      <td>0.0147241691924927</td>\n",
       "      <td>2.69</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>-1.35835406159823</td>\n",
       "      <td>-1.34016307473609</td>\n",
       "      <td>1.77320934263119</td>\n",
       "      <td>0.379779593034328</td>\n",
       "      <td>-0.503198133318193</td>\n",
       "      <td>1.80049938079263</td>\n",
       "      <td>0.791460956450422</td>\n",
       "      <td>0.247675786588991</td>\n",
       "      <td>-1.51465432260583</td>\n",
       "      <td>...</td>\n",
       "      <td>0.247998153469754</td>\n",
       "      <td>0.771679401917229</td>\n",
       "      <td>0.909412262347719</td>\n",
       "      <td>-0.689280956490685</td>\n",
       "      <td>-0.327641833735251</td>\n",
       "      <td>-0.139096571514147</td>\n",
       "      <td>-0.0553527940384261</td>\n",
       "      <td>-0.0597518405929204</td>\n",
       "      <td>378.66</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>-0.966271711572087</td>\n",
       "      <td>-0.185226008082898</td>\n",
       "      <td>1.79299333957872</td>\n",
       "      <td>-0.863291275036453</td>\n",
       "      <td>-0.0103088796030823</td>\n",
       "      <td>1.24720316752486</td>\n",
       "      <td>0.23760893977178</td>\n",
       "      <td>0.377435874652262</td>\n",
       "      <td>-1.38702406270197</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.108300452035545</td>\n",
       "      <td>0.00527359678253453</td>\n",
       "      <td>-0.190320518742841</td>\n",
       "      <td>-1.17557533186321</td>\n",
       "      <td>0.647376034602038</td>\n",
       "      <td>-0.221928844458407</td>\n",
       "      <td>0.0627228487293033</td>\n",
       "      <td>0.0614576285006353</td>\n",
       "      <td>123.5</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "      <td>-1.15823309349523</td>\n",
       "      <td>0.877736754848451</td>\n",
       "      <td>1.548717846511</td>\n",
       "      <td>0.403033933955121</td>\n",
       "      <td>-0.407193377311653</td>\n",
       "      <td>0.0959214624684256</td>\n",
       "      <td>0.592940745385545</td>\n",
       "      <td>-0.270532677192282</td>\n",
       "      <td>0.817739308235294</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.00943069713232919</td>\n",
       "      <td>0.79827849458971</td>\n",
       "      <td>-0.137458079619063</td>\n",
       "      <td>0.141266983824769</td>\n",
       "      <td>-0.206009587619756</td>\n",
       "      <td>0.502292224181569</td>\n",
       "      <td>0.219422229513348</td>\n",
       "      <td>0.215153147499206</td>\n",
       "      <td>69.99</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>6</td>\n",
       "      <td>2</td>\n",
       "      <td>-0.425965884412454</td>\n",
       "      <td>0.960523044882985</td>\n",
       "      <td>1.14110934232219</td>\n",
       "      <td>-0.168252079760302</td>\n",
       "      <td>0.42098688077219</td>\n",
       "      <td>-0.0297275516639742</td>\n",
       "      <td>0.476200948720027</td>\n",
       "      <td>0.260314333074874</td>\n",
       "      <td>-0.56867137571251</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.208253514656728</td>\n",
       "      <td>-0.559824796253248</td>\n",
       "      <td>-0.0263976679795373</td>\n",
       "      <td>-0.371426583174346</td>\n",
       "      <td>-0.232793816737034</td>\n",
       "      <td>0.105914779097957</td>\n",
       "      <td>0.253844224739337</td>\n",
       "      <td>0.0810802569229443</td>\n",
       "      <td>3.67</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>7</td>\n",
       "      <td>4</td>\n",
       "      <td>1.22965763450793</td>\n",
       "      <td>0.141003507049326</td>\n",
       "      <td>0.0453707735899449</td>\n",
       "      <td>1.20261273673594</td>\n",
       "      <td>0.191880988597645</td>\n",
       "      <td>0.272708122899098</td>\n",
       "      <td>-0.00515900288250983</td>\n",
       "      <td>0.0812129398830894</td>\n",
       "      <td>0.464959994783886</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.167716265815783</td>\n",
       "      <td>-0.270709726172363</td>\n",
       "      <td>-0.154103786809305</td>\n",
       "      <td>-0.780055415004671</td>\n",
       "      <td>0.75013693580659</td>\n",
       "      <td>-0.257236845917139</td>\n",
       "      <td>0.0345074297438413</td>\n",
       "      <td>0.00516776890624916</td>\n",
       "      <td>4.99</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>8</td>\n",
       "      <td>7</td>\n",
       "      <td>-0.644269442348146</td>\n",
       "      <td>1.41796354547385</td>\n",
       "      <td>1.0743803763556</td>\n",
       "      <td>-0.492199018495015</td>\n",
       "      <td>0.948934094764157</td>\n",
       "      <td>0.428118462833089</td>\n",
       "      <td>1.12063135838353</td>\n",
       "      <td>-3.80786423873589</td>\n",
       "      <td>0.615374730667027</td>\n",
       "      <td>...</td>\n",
       "      <td>1.94346533978412</td>\n",
       "      <td>-1.01545470979971</td>\n",
       "      <td>0.057503529867291</td>\n",
       "      <td>-0.649709005559993</td>\n",
       "      <td>-0.415266566234811</td>\n",
       "      <td>-0.0516342969262494</td>\n",
       "      <td>-1.20692108094258</td>\n",
       "      <td>-1.08533918832377</td>\n",
       "      <td>40.8</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>9</td>\n",
       "      <td>7</td>\n",
       "      <td>-0.89428608220282</td>\n",
       "      <td>0.286157196276544</td>\n",
       "      <td>-0.113192212729871</td>\n",
       "      <td>-0.271526130088604</td>\n",
       "      <td>2.6695986595986</td>\n",
       "      <td>3.72181806112751</td>\n",
       "      <td>0.370145127676916</td>\n",
       "      <td>0.851084443200905</td>\n",
       "      <td>-0.392047586798604</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.0734251001059225</td>\n",
       "      <td>-0.268091632235551</td>\n",
       "      <td>-0.204232669947878</td>\n",
       "      <td>1.0115918018785</td>\n",
       "      <td>0.373204680146282</td>\n",
       "      <td>-0.384157307702294</td>\n",
       "      <td>0.0117473564581996</td>\n",
       "      <td>0.14240432992147</td>\n",
       "      <td>93.2</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>10 rows × 31 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     0                   1                    2                   3   \\\n",
       "0  Time                  V1                   V2                  V3   \n",
       "1     0    -1.3598071336738  -0.0727811733098497    2.53634673796914   \n",
       "2     0    1.19185711131486     0.26615071205963    0.16648011335321   \n",
       "3     1   -1.35835406159823    -1.34016307473609    1.77320934263119   \n",
       "4     1  -0.966271711572087   -0.185226008082898    1.79299333957872   \n",
       "5     2   -1.15823309349523    0.877736754848451      1.548717846511   \n",
       "6     2  -0.425965884412454    0.960523044882985    1.14110934232219   \n",
       "7     4    1.22965763450793    0.141003507049326  0.0453707735899449   \n",
       "8     7  -0.644269442348146     1.41796354547385     1.0743803763556   \n",
       "9     7   -0.89428608220282    0.286157196276544  -0.113192212729871   \n",
       "\n",
       "                   4                    5                    6   \\\n",
       "0                  V4                   V5                   V6   \n",
       "1    1.37815522427443   -0.338320769942518    0.462387777762292   \n",
       "2   0.448154078460911   0.0600176492822243  -0.0823608088155687   \n",
       "3   0.379779593034328   -0.503198133318193     1.80049938079263   \n",
       "4  -0.863291275036453  -0.0103088796030823     1.24720316752486   \n",
       "5   0.403033933955121   -0.407193377311653   0.0959214624684256   \n",
       "6  -0.168252079760302     0.42098688077219  -0.0297275516639742   \n",
       "7    1.20261273673594    0.191880988597645    0.272708122899098   \n",
       "8  -0.492199018495015    0.948934094764157    0.428118462833089   \n",
       "9  -0.271526130088604      2.6695986595986     3.72181806112751   \n",
       "\n",
       "                     7                   8                   9   ...  \\\n",
       "0                    V7                  V8                  V9  ...   \n",
       "1     0.239598554061257  0.0986979012610507   0.363786969611213  ...   \n",
       "2   -0.0788029833323113  0.0851016549148104  -0.255425128109186  ...   \n",
       "3     0.791460956450422   0.247675786588991   -1.51465432260583  ...   \n",
       "4      0.23760893977178   0.377435874652262   -1.38702406270197  ...   \n",
       "5     0.592940745385545  -0.270532677192282   0.817739308235294  ...   \n",
       "6     0.476200948720027   0.260314333074874   -0.56867137571251  ...   \n",
       "7  -0.00515900288250983  0.0812129398830894   0.464959994783886  ...   \n",
       "8      1.12063135838353   -3.80786423873589   0.615374730667027  ...   \n",
       "9     0.370145127676916   0.851084443200905  -0.392047586798604  ...   \n",
       "\n",
       "                     21                   22                   23  \\\n",
       "0                   V21                  V22                  V23   \n",
       "1    -0.018306777944153    0.277837575558899   -0.110473910188767   \n",
       "2    -0.225775248033138   -0.638671952771851    0.101288021253234   \n",
       "3     0.247998153469754    0.771679401917229    0.909412262347719   \n",
       "4    -0.108300452035545  0.00527359678253453   -0.190320518742841   \n",
       "5  -0.00943069713232919     0.79827849458971   -0.137458079619063   \n",
       "6    -0.208253514656728   -0.559824796253248  -0.0263976679795373   \n",
       "7    -0.167716265815783   -0.270709726172363   -0.154103786809305   \n",
       "8      1.94346533978412    -1.01545470979971    0.057503529867291   \n",
       "9   -0.0734251001059225   -0.268091632235551   -0.204232669947878   \n",
       "\n",
       "                   24                  25                   26  \\\n",
       "0                 V24                 V25                  V26   \n",
       "1  0.0669280749146731   0.128539358273528   -0.189114843888824   \n",
       "2  -0.339846475529127   0.167170404418143    0.125894532368176   \n",
       "3  -0.689280956490685  -0.327641833735251   -0.139096571514147   \n",
       "4   -1.17557533186321   0.647376034602038   -0.221928844458407   \n",
       "5   0.141266983824769  -0.206009587619756    0.502292224181569   \n",
       "6  -0.371426583174346  -0.232793816737034    0.105914779097957   \n",
       "7  -0.780055415004671    0.75013693580659   -0.257236845917139   \n",
       "8  -0.649709005559993  -0.415266566234811  -0.0516342969262494   \n",
       "9     1.0115918018785   0.373204680146282   -0.384157307702294   \n",
       "\n",
       "                     27                   28      29     30  \n",
       "0                   V27                  V28  Amount  Class  \n",
       "1     0.133558376740387  -0.0210530534538215  149.62      0  \n",
       "2  -0.00898309914322813   0.0147241691924927    2.69      0  \n",
       "3   -0.0553527940384261  -0.0597518405929204  378.66      0  \n",
       "4    0.0627228487293033   0.0614576285006353   123.5      0  \n",
       "5     0.219422229513348    0.215153147499206   69.99      0  \n",
       "6     0.253844224739337   0.0810802569229443    3.67      0  \n",
       "7    0.0345074297438413  0.00516776890624916    4.99      0  \n",
       "8     -1.20692108094258    -1.08533918832377    40.8      0  \n",
       "9    0.0117473564581996     0.14240432992147    93.2      0  \n",
       "\n",
       "[10 rows x 31 columns]"
      ]
     },
     "execution_count": 1,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import pandas as pd \n",
    "df = pd.read_csv('creditcard.csv',header=None)\n",
    "df.head(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</th>\n",
       "      <th>...</th>\n",
       "      <th>21</th>\n",
       "      <th>22</th>\n",
       "      <th>23</th>\n",
       "      <th>24</th>\n",
       "      <th>25</th>\n",
       "      <th>26</th>\n",
       "      <th>27</th>\n",
       "      <th>28</th>\n",
       "      <th>29</th>\n",
       "      <th>30</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>count</td>\n",
       "      <td>284808</td>\n",
       "      <td>284808.000000</td>\n",
       "      <td>284808.000000</td>\n",
       "      <td>284808.000000</td>\n",
       "      <td>284808.000000</td>\n",
       "      <td>284808.000000</td>\n",
       "      <td>284808.000000</td>\n",
       "      <td>284808.000000</td>\n",
       "      <td>284808.000000</td>\n",
       "      <td>284808.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>284808.000000</td>\n",
       "      <td>284808.000000</td>\n",
       "      <td>284808.000000</td>\n",
       "      <td>284808.000000</td>\n",
       "      <td>284808.000000</td>\n",
       "      <td>284808.000000</td>\n",
       "      <td>284808.000000</td>\n",
       "      <td>284808.000000</td>\n",
       "      <td>284808.0</td>\n",
       "      <td>284808</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>unique</td>\n",
       "      <td>124593</td>\n",
       "      <td>276177.000000</td>\n",
       "      <td>276177.000000</td>\n",
       "      <td>276177.000000</td>\n",
       "      <td>276177.000000</td>\n",
       "      <td>276177.000000</td>\n",
       "      <td>276177.000000</td>\n",
       "      <td>276177.000000</td>\n",
       "      <td>276177.000000</td>\n",
       "      <td>276177.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>276177.000000</td>\n",
       "      <td>276177.000000</td>\n",
       "      <td>276177.000000</td>\n",
       "      <td>276177.000000</td>\n",
       "      <td>276177.000000</td>\n",
       "      <td>276177.000000</td>\n",
       "      <td>276177.000000</td>\n",
       "      <td>276177.000000</td>\n",
       "      <td>40135.0</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>top</td>\n",
       "      <td>163152</td>\n",
       "      <td>2.055797</td>\n",
       "      <td>-0.326668</td>\n",
       "      <td>-2.752041</td>\n",
       "      <td>-0.842316</td>\n",
       "      <td>2.463072</td>\n",
       "      <td>3.173856</td>\n",
       "      <td>-0.432126</td>\n",
       "      <td>0.727706</td>\n",
       "      <td>0.608606</td>\n",
       "      <td>...</td>\n",
       "      <td>0.269765</td>\n",
       "      <td>0.844627</td>\n",
       "      <td>0.020675</td>\n",
       "      <td>0.726212</td>\n",
       "      <td>0.366624</td>\n",
       "      <td>-0.398828</td>\n",
       "      <td>0.027735</td>\n",
       "      <td>-0.060282</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>freq</td>\n",
       "      <td>36</td>\n",
       "      <td>77.000000</td>\n",
       "      <td>77.000000</td>\n",
       "      <td>77.000000</td>\n",
       "      <td>77.000000</td>\n",
       "      <td>77.000000</td>\n",
       "      <td>77.000000</td>\n",
       "      <td>77.000000</td>\n",
       "      <td>77.000000</td>\n",
       "      <td>77.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>77.000000</td>\n",
       "      <td>77.000000</td>\n",
       "      <td>77.000000</td>\n",
       "      <td>77.000000</td>\n",
       "      <td>77.000000</td>\n",
       "      <td>77.000000</td>\n",
       "      <td>77.000000</td>\n",
       "      <td>77.000000</td>\n",
       "      <td>12295.0</td>\n",
       "      <td>251650</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>4 rows × 31 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "            0              1              2              3              4   \\\n",
       "count   284808  284808.000000  284808.000000  284808.000000  284808.000000   \n",
       "unique  124593  276177.000000  276177.000000  276177.000000  276177.000000   \n",
       "top     163152       2.055797      -0.326668      -2.752041      -0.842316   \n",
       "freq        36      77.000000      77.000000      77.000000      77.000000   \n",
       "\n",
       "                   5              6              7              8   \\\n",
       "count   284808.000000  284808.000000  284808.000000  284808.000000   \n",
       "unique  276177.000000  276177.000000  276177.000000  276177.000000   \n",
       "top          2.463072       3.173856      -0.432126       0.727706   \n",
       "freq        77.000000      77.000000      77.000000      77.000000   \n",
       "\n",
       "                   9   ...             21             22             23  \\\n",
       "count   284808.000000  ...  284808.000000  284808.000000  284808.000000   \n",
       "unique  276177.000000  ...  276177.000000  276177.000000  276177.000000   \n",
       "top          0.608606  ...       0.269765       0.844627       0.020675   \n",
       "freq        77.000000  ...      77.000000      77.000000      77.000000   \n",
       "\n",
       "                   24             25             26             27  \\\n",
       "count   284808.000000  284808.000000  284808.000000  284808.000000   \n",
       "unique  276177.000000  276177.000000  276177.000000  276177.000000   \n",
       "top          0.726212       0.366624      -0.398828       0.027735   \n",
       "freq        77.000000      77.000000      77.000000      77.000000   \n",
       "\n",
       "                   28        29      30  \n",
       "count   284808.000000  284808.0  284808  \n",
       "unique  276177.000000   40135.0       5  \n",
       "top         -0.060282       1.0       0  \n",
       "freq        77.000000   12295.0  251650  \n",
       "\n",
       "[4 rows x 31 columns]"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "clas_weight manual :{0.0: 1.0, 1.0: 1.9965415084816431}\n",
      "clas_weight Scikit-learn :{0.0: 0.5008661206149896, 1.0: 289.14340101522845}\n",
      "{0.0: 1.0, 1.0: 1.9965415084816431}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\sklearn\\utils\\validation.py:71: FutureWarning: Pass classes=[0. 1.], y=[0. 0. 0. ... 0. 0. 0.] as keyword args. From version 0.25 passing these as positional arguments will result in an error\n",
      "  FutureWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 227845 samples, validate on 56962 samples\n",
      "Epoch 1/60\n",
      "227845/227845 [==============================] - 9s 39us/step - loss: 0.0086 - accuracy: 0.9991 - val_loss: 0.0036 - val_accuracy: 0.9994\n",
      "Epoch 2/60\n",
      "227845/227845 [==============================] - 9s 40us/step - loss: 0.0067 - accuracy: 0.9993 - val_loss: 0.0053 - val_accuracy: 0.9992\n",
      "Epoch 3/60\n",
      "227845/227845 [==============================] - 10s 42us/step - loss: 0.0067 - accuracy: 0.9993 - val_loss: 0.0037 - val_accuracy: 0.9993\n",
      "Epoch 4/60\n",
      "227845/227845 [==============================] - 9s 39us/step - loss: 0.0065 - accuracy: 0.9992 - val_loss: 0.0034 - val_accuracy: 0.9995\n",
      "Epoch 5/60\n",
      "227845/227845 [==============================] - 10s 45us/step - loss: 0.0069 - accuracy: 0.9993 - val_loss: 0.0037 - val_accuracy: 0.9993\n",
      "Epoch 6/60\n",
      "227845/227845 [==============================] - 10s 42us/step - loss: 0.0063 - accuracy: 0.9992 - val_loss: 0.0035 - val_accuracy: 0.9993\n",
      "Epoch 7/60\n",
      "227845/227845 [==============================] - 10s 45us/step - loss: 0.0071 - accuracy: 0.9993 - val_loss: 0.0031 - val_accuracy: 0.9995\n",
      "Epoch 8/60\n",
      "227845/227845 [==============================] - 9s 39us/step - loss: 0.0064 - accuracy: 0.9993 - val_loss: 0.0040 - val_accuracy: 0.9991\n",
      "Epoch 9/60\n",
      "227845/227845 [==============================] - 9s 38us/step - loss: 0.0064 - accuracy: 0.9993 - val_loss: 0.0033 - val_accuracy: 0.9995\n",
      "Epoch 10/60\n",
      "227845/227845 [==============================] - 9s 40us/step - loss: 0.0061 - accuracy: 0.9994 - val_loss: 0.0032 - val_accuracy: 0.9995\n",
      "Epoch 11/60\n",
      "227845/227845 [==============================] - 11s 50us/step - loss: 0.0065 - accuracy: 0.9994 - val_loss: 0.0037 - val_accuracy: 0.9995\n",
      "Epoch 12/60\n",
      "227845/227845 [==============================] - 9s 38us/step - loss: 0.0063 - accuracy: 0.9993 - val_loss: 0.0033 - val_accuracy: 0.9995\n",
      "Epoch 13/60\n",
      "227845/227845 [==============================] - 9s 39us/step - loss: 0.0064 - accuracy: 0.9993 - val_loss: 0.0035 - val_accuracy: 0.9992\n",
      "Epoch 14/60\n",
      "227845/227845 [==============================] - 9s 39us/step - loss: 0.0062 - accuracy: 0.9992 - val_loss: 0.0038 - val_accuracy: 0.9992\n",
      "Epoch 15/60\n",
      "227845/227845 [==============================] - 8s 37us/step - loss: 0.0060 - accuracy: 0.9993 - val_loss: 0.0032 - val_accuracy: 0.9995\n",
      "Epoch 16/60\n",
      "227845/227845 [==============================] - 8s 37us/step - loss: 0.0062 - accuracy: 0.9994 - val_loss: 0.0033 - val_accuracy: 0.9995\n",
      "Epoch 17/60\n",
      "227845/227845 [==============================] - 8s 37us/step - loss: 0.0062 - accuracy: 0.9993 - val_loss: 0.0035 - val_accuracy: 0.9994\n",
      "Epoch 18/60\n",
      "227845/227845 [==============================] - 8s 37us/step - loss: 0.0063 - accuracy: 0.9993 - val_loss: 0.0031 - val_accuracy: 0.9995\n",
      "Epoch 19/60\n",
      "227845/227845 [==============================] - 8s 37us/step - loss: 0.0071 - accuracy: 0.9993 - val_loss: 0.0035 - val_accuracy: 0.9994\n",
      "Epoch 20/60\n",
      "227845/227845 [==============================] - 8s 37us/step - loss: 0.0062 - accuracy: 0.9993 - val_loss: 0.0034 - val_accuracy: 0.9995\n",
      "Epoch 21/60\n",
      "227845/227845 [==============================] - 9s 39us/step - loss: 0.0068 - accuracy: 0.9993 - val_loss: 0.0038 - val_accuracy: 0.9992\n",
      "Epoch 22/60\n",
      "227845/227845 [==============================] - 9s 38us/step - loss: 0.0064 - accuracy: 0.9992 - val_loss: 0.0034 - val_accuracy: 0.9995\n",
      "Epoch 23/60\n",
      "227845/227845 [==============================] - 9s 37us/step - loss: 0.0063 - accuracy: 0.9993 - val_loss: 0.0035 - val_accuracy: 0.9994\n",
      "Epoch 24/60\n",
      "227845/227845 [==============================] - 9s 38us/step - loss: 0.0064 - accuracy: 0.9993 - val_loss: 0.0032 - val_accuracy: 0.9995\n",
      "Epoch 25/60\n",
      "227845/227845 [==============================] - 9s 38us/step - loss: 0.0064 - accuracy: 0.9994 - val_loss: 0.0042 - val_accuracy: 0.9992\n",
      "Epoch 26/60\n",
      "227845/227845 [==============================] - 9s 37us/step - loss: 0.0063 - accuracy: 0.9994 - val_loss: 0.0039 - val_accuracy: 0.9992\n",
      "Epoch 27/60\n",
      "227845/227845 [==============================] - 8s 37us/step - loss: 0.0063 - accuracy: 0.9993 - val_loss: 0.0033 - val_accuracy: 0.9995\n",
      "Epoch 28/60\n",
      "227845/227845 [==============================] - 9s 38us/step - loss: 0.0066 - accuracy: 0.9994 - val_loss: 0.0041 - val_accuracy: 0.9993\n",
      "Epoch 29/60\n",
      "227845/227845 [==============================] - 9s 39us/step - loss: 0.0066 - accuracy: 0.9994 - val_loss: 0.0034 - val_accuracy: 0.9991\n",
      "Epoch 30/60\n",
      "227845/227845 [==============================] - 8s 37us/step - loss: 0.0063 - accuracy: 0.9993 - val_loss: 0.0036 - val_accuracy: 0.9994\n",
      "Epoch 31/60\n",
      "227845/227845 [==============================] - 9s 38us/step - loss: 0.0065 - accuracy: 0.9993 - val_loss: 0.0038 - val_accuracy: 0.9992\n",
      "Epoch 32/60\n",
      "227845/227845 [==============================] - 9s 37us/step - loss: 0.0061 - accuracy: 0.9993 - val_loss: 0.0039 - val_accuracy: 0.9992\n",
      "Epoch 33/60\n",
      "227845/227845 [==============================] - 9s 38us/step - loss: 0.0061 - accuracy: 0.9993 - val_loss: 0.0038 - val_accuracy: 0.9993\n",
      "Epoch 34/60\n",
      "227845/227845 [==============================] - 9s 38us/step - loss: 0.0062 - accuracy: 0.9993 - val_loss: 0.0033 - val_accuracy: 0.9995\n",
      "Epoch 35/60\n",
      "227845/227845 [==============================] - 9s 38us/step - loss: 0.0062 - accuracy: 0.9994 - val_loss: 0.0033 - val_accuracy: 0.9995\n",
      "Epoch 36/60\n",
      "227845/227845 [==============================] - 9s 37us/step - loss: 0.0063 - accuracy: 0.9993 - val_loss: 0.0033 - val_accuracy: 0.9995\n",
      "Epoch 37/60\n",
      "227845/227845 [==============================] - 9s 40us/step - loss: 0.0062 - accuracy: 0.9994 - val_loss: 0.0034 - val_accuracy: 0.9995\n",
      "Epoch 38/60\n",
      "227845/227845 [==============================] - 8s 37us/step - loss: 0.0062 - accuracy: 0.9993 - val_loss: 0.0035 - val_accuracy: 0.9995\n",
      "Epoch 39/60\n",
      "227845/227845 [==============================] - 8s 37us/step - loss: 0.0061 - accuracy: 0.9994 - val_loss: 0.0037 - val_accuracy: 0.9993\n",
      "Epoch 40/60\n",
      "227845/227845 [==============================] - 8s 37us/step - loss: 0.0060 - accuracy: 0.9993 - val_loss: 0.0036 - val_accuracy: 0.9993\n",
      "Epoch 41/60\n",
      "227845/227845 [==============================] - 9s 38us/step - loss: 0.0064 - accuracy: 0.9993 - val_loss: 0.0042 - val_accuracy: 0.9994\n",
      "Epoch 42/60\n",
      "227845/227845 [==============================] - 9s 38us/step - loss: 0.0082 - accuracy: 0.9993 - val_loss: 0.0035 - val_accuracy: 0.9993\n",
      "Epoch 43/60\n",
      "227845/227845 [==============================] - 9s 38us/step - loss: 0.0064 - accuracy: 0.9993 - val_loss: 0.0034 - val_accuracy: 0.9995\n",
      "Epoch 44/60\n",
      "227845/227845 [==============================] - 11s 50us/step - loss: 0.0063 - accuracy: 0.9994 - val_loss: 0.0038 - val_accuracy: 0.9993\n",
      "Epoch 45/60\n",
      "227845/227845 [==============================] - 9s 39us/step - loss: 0.0063 - accuracy: 0.9993 - val_loss: 0.0044 - val_accuracy: 0.9989\n",
      "Epoch 46/60\n",
      "227845/227845 [==============================] - 9s 38us/step - loss: 0.0063 - accuracy: 0.9993 - val_loss: 0.0034 - val_accuracy: 0.9995\n",
      "Epoch 47/60\n",
      "227845/227845 [==============================] - 9s 37us/step - loss: 0.0061 - accuracy: 0.9994 - val_loss: 0.0038 - val_accuracy: 0.9994\n",
      "Epoch 48/60\n",
      "227845/227845 [==============================] - 11s 48us/step - loss: 0.0068 - accuracy: 0.9994 - val_loss: 0.0034 - val_accuracy: 0.9995\n",
      "Epoch 49/60\n",
      "227845/227845 [==============================] - 9s 40us/step - loss: 0.0062 - accuracy: 0.9992 - val_loss: 0.0038 - val_accuracy: 0.9992\n",
      "Epoch 50/60\n",
      "227845/227845 [==============================] - 9s 39us/step - loss: 0.0064 - accuracy: 0.9993 - val_loss: 0.0035 - val_accuracy: 0.9995\n",
      "Epoch 51/60\n",
      "227845/227845 [==============================] - 9s 40us/step - loss: 0.0063 - accuracy: 0.9994 - val_loss: 0.0032 - val_accuracy: 0.9995\n",
      "Epoch 52/60\n",
      "227845/227845 [==============================] - 9s 39us/step - loss: 0.0059 - accuracy: 0.9994 - val_loss: 0.0031 - val_accuracy: 0.9995\n",
      "Epoch 53/60\n",
      "227845/227845 [==============================] - 9s 39us/step - loss: 0.0067 - accuracy: 0.9994 - val_loss: 0.0035 - val_accuracy: 0.9993\n",
      "Epoch 54/60\n",
      "227845/227845 [==============================] - 9s 38us/step - loss: 0.0067 - accuracy: 0.9994 - val_loss: 0.0031 - val_accuracy: 0.9996\n",
      "Epoch 55/60\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "227845/227845 [==============================] - 9s 41us/step - loss: 0.0061 - accuracy: 0.9994 - val_loss: 0.0035 - val_accuracy: 0.9995\n",
      "Epoch 56/60\n",
      "227845/227845 [==============================] - 9s 41us/step - loss: 0.0060 - accuracy: 0.9993 - val_loss: 0.0033 - val_accuracy: 0.9995\n",
      "Epoch 57/60\n",
      "227845/227845 [==============================] - 9s 39us/step - loss: 0.0058 - accuracy: 0.9994 - val_loss: 0.0031 - val_accuracy: 0.9995\n",
      "Epoch 58/60\n",
      "227845/227845 [==============================] - 9s 39us/step - loss: 0.0060 - accuracy: 0.9994 - val_loss: 0.0033 - val_accuracy: 0.9995\n",
      "Epoch 59/60\n",
      "227845/227845 [==============================] - 9s 40us/step - loss: 0.0058 - accuracy: 0.9993 - val_loss: 0.0034 - val_accuracy: 0.9994\n",
      "Epoch 60/60\n",
      "227845/227845 [==============================] - 9s 39us/step - loss: 0.0063 - accuracy: 0.9993 - val_loss: 0.0032 - val_accuracy: 0.9996\n",
      "dict_keys(['val_loss', 'val_accuracy', 'loss', 'accuracy'])\n",
      "       reconstruction_error    true_class\n",
      "count          56962.000000  56962.000000\n",
      "mean               0.158122      0.001720\n",
      "std                0.040540      0.041443\n",
      "min                0.056955      0.000000\n",
      "25%                0.151036      0.000000\n",
      "50%                0.155742      0.000000\n",
      "75%                0.161448      0.000000\n",
      "max                1.502492      1.000000\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAfEAAAHwCAYAAAC2blbYAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjMsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+AADFEAAAgAElEQVR4nO3deZgdVZ3/8fc3G4uCURIEshBQFhEFNTA4OMCMioAKogyCouKgiIrKz11klFFnxg0XRlxQFEUBwRkxSpRRQMBRkKCAEI3GCCQssgQiSCDb9/fHuW3fdDrpm6Srb5/u9+t56rlVdavrfm9l+fQ5daoqMhNJklSfMd0uQJIkbRhDXJKkShnikiRVyhCXJKlShrgkSZUyxCVJqpQhLlUuIg6IiEVty7dExPO6WdNgiIh/iIh5HWx3ckR8ZShqkoYbQ1waRK0AXRoRD0XEXRFxdkQ8ttt1DabWLw2rWt/xwYiYFxGvHezPycyrMnOXDrb7j8x83WB/vlQDQ1wafC/OzMcCewLPAN7X5XqacEfrO24JvAf4ckTs1nejiBg35JVJo4ghLjUkM+8CLqGEOQARsUlEfDIibouIP0fEFyNis7b3D4uI6yPiLxHxx4g4qLX+tRHx21bLd0FEvGF964mIfVq9A2Pb1h0eETe25veOiDmtz/5zRHyqg++YmXkRcD+wW0TMiIiMiOMi4jbgsrbP/nlEPBARN0TEAW01PCEivhYRd0TE/RFxUWt939ME74mI29ta/89trT81Ir7Ztt2hEXFz67N+GhFPaXvvloh4Z0TcGBFLIuLbEbHp+h5LabgwxKWGRMRU4GBgftvqjwE7U4L9ycAU4AOt7fcGvgG8C5gI7Afc0vq5u4EXUVq+rwU+HRHPXJ96MvNq4K/AP7WtfgVwbmv+s8BnM3NL4EnABR18xzERcXir3t+0vbU/8BTgBRExBbgY+AjwBOCdwH9HxOTWtucAmwNPBbYGPt3P5+wCnAjslZlbAC+g99i0b7czcB5wEjAZmA18PyImtG12JHAQsAPwdODYgb6nNFwZ4tLguygiHgQWUsL3gwAREcDrgf+XmYsz80HgP4CjWj93HPDVzPxxZq7KzNsz83cAmXlxZv6x1fK9Avhf4B82oLbzgKNb9WwBHNJaB7AceHJETMrMh1qhvzbbRcQDwL2t7/eqzGwfhHZqZv41M5cCxwCzM3N263v9GJgDHBIR21J+0TkhM+/PzOWt79fXSmATSmt/fGbekpl/7Ge7lwMXt47hcuCTwGbA37dtc3pm3pGZi4Hv09ZTItXGEJcG30tarcUDgF2BSa31kyktzutaXb0PAD9qrQeYBvQXTETEwRFxdUQsbv3cIW37XR/nAi+NiE2AlwK/ysxbW+8dR+kl+F1EXBsRL1rHfu7IzImZ+YTM3DMzz+/z/sK2+e2Bf+75zq36nwNsS/nOizPz/nUVnZnzKa3rU4G7I+L8iNiun023A25t+7lVrVqmtG1zV9v8w8CIGnio0cUQlxrSalGeTWkNQmm1LgWe2grAiZn5uNYAMShh86S++2kF7n+39vPEzJxI6SaODahpLiXkDmb1rnQy8w+ZeTSlS/tjwHci4jHr+xk9u2ubXwic0/adJ2bmYzLzo633nhAREzuo/dzMfA7ll4Js1djXHa33gb/1fkwDbt/A7yENa4a41KzPAM+PiD1brcIvU85nbw0QEVMi4gWtbc8CXhsRz22da54SEbsCEyhdyfcAKyLiYODAjajpXOCtlHPuF/asjIhjImJyq84HWqtXbsTn9Pgm8OKIeEFEjI2ITVuD1qZm5p3AD4HPR8TjI2J8ROzXdwcRsUtE/FPrF5pHKL8M9VfbBcALW8dwPPAO4FHg54PwPaRhxxCXGpSZ91AGq/1ra9V7KAPdro6IvwA/AXZpbftLWoPWgCXAFcD2rXPnb6UE1P2UFvSsjSjrPEpX/2WZeW/b+oOAmyPiIcogt6My85GN+BwAMnMhcBhwMuUXkYWUwXs9//+8inI+/neUMQQn9bObTYCPUnoz7qL0Fpzcz2fNo5yD/6/Wti+mXPK3bGO/hzQcRWYOvJUkSRp2bIlLklQpQ1ySpEoZ4pIkVcoQlySpUoa4JEmVqu4JQ5MmTcoZM2Z0uwxJkobEddddd29mTu7vvepCfMaMGcyZM6fbZUiSNCQi4ta1vWd3uiRJlTLEJUmqlCEuSVKlDHFJkipliEuSVClDXJKkShnikiRVyhCXJKlShrgkSZUyxCVJqpQhLklSpQxxSZIqZYhLklQpQ1ySpEo1FuIR8dWIuDsiblrL+xERp0fE/Ii4MSKe2VQtkiSNRE22xM8GDlrH+wcDO7Wm44EvNFiLJEkjzrimdpyZV0bEjHVschjwjcxM4OqImBgR22bmnU3V1Nejj8LSpUP1aUPrMY+B8eO7XYUkqUmNhXgHpgAL25YXtdYNWYh/61tw3HFD9WlD66lPhZv6PZEhSRopuhni0c+67HfDiOMpXe5Mnz590Ar4u7+DT3960HY3bFx0EVx3XberkCQ1rZshvgiY1rY8Fbijvw0z80zgTICZM2f2G/Qb4qlPLdNIs3ChIS5Jo0E3LzGbBby6NUp9H2DJUJ4PlySpdo21xCPiPOAAYFJELAI+CIwHyMwvArOBQ4D5wMPAa5uqRZKkkajJ0elHD/B+Am9u6vMlSRrpvGObJEmVMsQlSaqUIS5JUqUMcUmSKmWIS5JUKUNckqRKGeKSJFXKEJckqVKGuCRJlTLEJUmqlCEuSVKlDHFJkipliEuSVClDXJKkShnikiRVyhCXJKlShrgkSZUyxCVJqpQhLklSpQxxSZIqZYiPcMuXw7e+BR/8IGR2uxpJ0mAa1+0C1IyVK+FTn4LPfAYWLizr3vQmeOITu1uXJGnw2BIfoZYuhXe8A3bcEV75ym5XI0lqgiE+Ar30pfCGN8Avfwk//Snsu2+3K5IkNcHu9BFo330NbkkaDWyJS5LWasUKuPNOmDcPVq3qdjXqy5a4JI0ymfDAA3DXXQNP99zTe2XLhRfCEUd0t3atzhCXpBGip9V8++3ldV3hvGzZmj8/YQJss02ZZsyAffYp8xMmwCmnwP33D/lX0gAMcUmqwIMPlnC+/XZYtKh3vn2666417wcRAZMn94bzrrv2zvedJk4s2/d1++0lxDX8GOKS1EWrVsHdd689mHvWP/jgmj/7+MfDlCll2mOP3vkpU2Dbbcs0eTKM83/6Ecs/Wklq0COPwG23lenWW9ecX7hwza7tceNKAE+ZAk99Khx4YJmfOrU3pLfbDjbfvDvfScOHIS5JGygTFi/uDeT+Xu++e/WfGTOmBPD06bD33mWg2PTpq7eit94axo7tzndSXQxxSVqHZctKGC9Y0P/0l7+svv1mm8H225dg3nPP8tqzvP32JaTHj+/Od9HIY4hLGtUy4b77SiD/8Y9rhvSiRatfH73JJuV2xjvsAM95TnndYYfekN5qq/4Hh0lNMMQljQoPPAC///2a0/z5aw4a22abEtT77Vden/Sk8rrjjuW9Md4mS8OEIS5pxHj00dKanjdvzbBuPzc9ZkxpPe+8c2lNt4f0Djs4YEz1MMQlVSWzBPLcuWVqD+xbb12963ubbUpQH3poed1ll/K6447lBiZS7QxxScNSJvz5zyWob7559df77uvd7rGPLeG8zz7w6leXkO6Zttyye/VLQ8EQl9R1ixfDDTfATTeVoO4J68WLe7eZOLFcM/2yl8Fuu5X53XYr11M7kEyjlSEuacisXAl/+EMJ7BtugBtvLK+LFvVu8/jHl4D+539ePay32cawlvoyxCU1YsmS3pDumW66CZYuLe+PG1fu473//uWWoXvsAU97mmEtrQ9DXNJGe/BB+PWvYc6c3ukPf+h9f6utSkifcEJ5ffrTS+t6k026V7M0EhjiktbLX/8K11+/emDPm9f79Kxp02DmTHjNa+AZzyihvd12tq6lJhji0gZasgQuvhgeegiOP77b1TRj5coyyOznP4drrimBPXdu72Vc221XAvsVryivz3pWue+3pKFhiEvr4c474Xvfg+9+Fy6/HJYvL+tf8YpyqVPtHnighPXPf94b3D13M9t6a9hrrzI6/FnPKtN223W3Xmm0M8SlAfz+93DRRSW4r766rHvyk+Gkk8r1yl/9am9Xck0yy3f7xS96Q3vu3LJ+zJhy3vpVr4K///syzZhhl7g03BjiUj/+9Cc4/3w499wyohpKy/PDH4bDDy+DsiLgtNO6W+f6WLGinMu+4ooy/fznvTdNmTgRnv1sOOqoEth77QVbbNHdeiUNzBCXWv78Z7jgAjjvvNI6hRJon/lMCe7p07tb3/pasQJ+9Sv46U9LaP/sZ72Pzdxpp3Ir0n33Ld9xl118qIdUI0Nco9qSJaWb/Nxz4dJLy4Ctpz0N/vM/S6t0xoxuV9i5VavgN7+Bn/ykTD/7WRl0B+V67KOPLtdk77+/57KlkcIQ16iTCVdeCV/5CnznO/DII+XJVe99bwm63XfvdoWdW7QIfvzjMl16ae+TunbdtZzPPuCA8jjNbbbpapmSGmKIa9S46y74+tfhrLPKjUi23BKOPbZMe+9dx6Cthx8uo+J/9KMS3PPmlfVPfCI8//lleu5zYerU7tYpaWgY4hrRVqyASy4pre7vf79c9/wP/wCnnAJHHFHHc6MXLIDZs8s16ZdfXp6ZvfnmpVv8+ONLcO++ex2/hEgaXIa4RqR774UvfQm+8AW4/fZyjfPb3w7HHVcGcQ1ny5bBVVeV4J49G373u7J+p53KbUtf+MLSRe4tSyUZ4hpR5s4to8nPOaec637+8+H00+FFL4IJE7pd3do99BD88IdlkN3FF5dR5BMmlHPaJ5wAhxxSQlyS2hniql5m6TL/9Kfhf/8XNt20DOp629vKYyyHq3vvLV383/1uqfvRR2HSpNLNf+ih5dz2SLgLnKTmGOKq1sMPlxb3Zz8Lv/0tbLstfOQj8IY3lDAcju66Cy68EP7nf8oI+VWryvXnJ5xQrkXfd9/yiE5J6oT/Xag6f/lLOdd92mlwzz3wzGeWMD/yyOHZZb54cQnt884rN15Ztar0EJx8cgnuZzzDQWmSNowhrmosXlzOb59+Otx/Pxx4YAnC/fYbfiH40EPlQSnnn1+6+pcvL+e0Tzml3ETmKU/pdoWSRgJDXMPekiVlsNqnPlVa4S95SQnvvfbqdmWrW768XL99zjnwgx/A0qXl2dpve1u5iYwtbkmDzRDXsPXQQ/C5z8HHP15a3i99KZx6arkt6nDz7neXLvO774bJk+Ff/qX3YSLek1xSUwxxDTsrVpSbs5x6ankoyQtfCB/6UDn3Pdz0nIM/6yx48YvhNa+Bgw+G8eO7W5ek0cEQ17CRCbNmwXveU24n+pznlMuvnv3sble2dsccU24k87znwVZbdbsaSaONHX0aFm64odzY5CUvKcsXXVQuwRrOAQ7w+MfDy19ugEvqDkNcXbV4MbzpTaWrfO5c+Pzny+M0DzvMQWCSNBC709UVq1bBl79cRpkvWQJvfjP827+Vlq0kqTOGuDqyalVpGQ9G6/imm8rTt37xi9KFfvrpw3PEuSQNd3ana0C/+Q3svDO8970bt5+lS+H97y/XS//hD+V66ssuM8AlaUPZEtc6zZoFr3xluWb7tts2fD9XXVWunZ4/v1yG9clPDt/7m0tSLWyJq1+Z8NGPltHiu+4K22yzYft55BF45zth//1h5Ur4yU/g7LMNcEkaDIa41rBiBRx7LLzvfeWuY1deCVtuuf77mTOnjDo/7bTyZLEbbyyP15QkDY5GQzwiDoqIeRExPyLWOKMaEdMj4vKI+HVE3BgRhzRZjwa2YkW5gck3vlHukvatb8Fmm63fPlauhH//d9hnn3Kv8x/9qDx1zGdjS9LgauyceESMBc4Ang8sAq6NiFmZObdts1OACzLzCxGxGzAbmNFUTVq3lSvL+epvfxs+8YnSDb6+7ryz/BJw2WWlFf/5z3vZmCQ1pcmW+N7A/MxckJnLgPOBw/psk0BPR+3jgDsarEfrsHJl6UI/99xyLnxDAvxHP4I99iiXjp11VtmXAS5JzWkyxKcAC9uWF7XWtTsVOCYiFlFa4W9psB6txcqVZeT4N79ZusHf8571//n3v788+GObbeC668r+vOOaJDWryRDv77/w7LN8NHB2Zk4FDgHOiYg1aoqI4yNiTkTMueeeexoodfRatQpe97rec+Ann7x+P794cXnK2H/8B7z+9XDNNfCUpzRTqyRpdU2G+CJgWtvyVNbsLj8OuAAgM38BbAqscfFRZp6ZmTMzc+bkyZMbKnf0WbWq3Dnt7LPhgx+Ef/3X9fv5G2+EvfaCyy+HM88s0/oOgpMkbbgmQ/xaYKeI2CEiJgBHAbP6bHMb8FyAiHgKJcRtag+BVavgjW8s565POaWE+Pq48MLyhLFHHoErriitcEnS0GosxDNzBXAicAnwW8oo9Jsj4kMRcWhrs3cAr4+IG4DzgGMzs2+XuwZZJpx4Ymk5v+99pRu90/PXmfCxj8GRR8Kee5bz3/vs02y9kqT+NXrb1cycTRmw1r7uA23zc4F9m6xBq8uEt7ylXLf97neXgWydBvjy5eWmLV/+crl87Gtfg003bbZeSdLaee/0USQTTjoJzjgD3vGOcinZ+owg/5//Kfs4+WT48IdhjPf7k6SuMsRHkXe/uzw57KSTys1c1ifAx4wp05e+BMcd11yNkqTOGeKjyDnnlK70T31q/a/h/sQnyo1b9vXkhyQNG4b4KDB+fHl905vgs5/dsJuwvOhFg1uTJGnjGeKjwBFHlFb04Yd7FzVJGkkM8VFg4kR42cu6XYUkabA5vliSpEoZ4pIkVcoQlySpUoa4JEmVMsQlSaqUIS5JUqUMcUmSKmWIS5JUKUNckqRKGeKSJFXKEJckqVKGuCRJlTLEJUmqlCEuSVKlDHFJkipliEuSVClDXJKkShnikiRVyhCXJKlShrgkSZUyxCVJqpQhLklSpQxxSZIqZYhLklQpQ1ySpEoZ4pIkVcoQlySpUoa4JEmVMsQlSaqUIS5JUqUMcUmSKmWIS5JUKUNckqRKGeKSJFXKEJckqVKGuCRJlTLEJUmqlCEuSVKlDHFJkipliEuSVClDXJKkShnikiRVyhCXJKlShrgkSZUyxCVJqpQhLklSpQxxSZIqZYhLklQpQ1ySpEoZ4pIkVcoQlySpUoa4JEmVMsQlSaqUIS5JUqUMcUmSKmWIS5JUKUNckqRKGeKSJFXKEJckqVKGuCRJlTLEJUmqlCEuSVKlDHFJkipliEuSVKlGQzwiDoqIeRExPyLeu5ZtjoyIuRFxc0Sc22Q9kiSNJOOa2nFEjAXOAJ4PLAKujYhZmTm3bZudgPcB+2bm/RGxdVP1SJI00jTZEt8bmJ+ZCzJzGXA+cFifbV4PnJGZ9wNk5t0N1iNJ0ojSZIhPARa2LS9qrWu3M7BzRPxfRFwdEQc1WI8kSSNKY93pQPSzLvv5/J2AA4CpwFURsXtmPrDajiKOB44HmD59+uBXKklShZpsiS8CprUtTwXu6Geb72Xm8sz8EzCPEuqrycwzM3NmZs6cPHlyYwVLklSTJkP8WmCniNghIiYARwGz+mxzEfCPABExidK9vqDBmiRJGjEaC/HMXAGcCFwC/Ba4IDNvjogPRcShrc0uAe6LiLnA5cC7MvO+pmqSJGkkafKcOJk5G5jdZ90H2uYTeHtrkiRJ68E7tkmSVClDXJKkShnikiRVyhCXJKlShrgkSZUyxCVJqpQhLklSpQxxSZIqZYhLklQpQ1ySpEoZ4pIkVcoQlySpUoa4JEmVMsQlSapUx48ijYgpwPbtP5OZVzZRlCRJGlhHIR4RHwNeDswFVrZWJ2CIS5LUJZ22xF8C7JKZjzZZjCRJ6lyn58QXAOObLESSJK2fTlviDwPXR8SlwN9a45n51kaqkiRJA+o0xGe1JkmSNEx0FOKZ+fWImADs3Fo1LzOXN1eWJEkaSKej0w8Avg7cAgQwLSJe4yVmkiR1T6fd6acBB2bmPICI2Bk4D3hWU4VJkqR163R0+vieAAfIzN/jaHVJkrqq05b4nIg4CzintfxK4LpmSpIkSZ3oNMTfCLwZeCvlnPiVwOebKkqSJA2s09HpjwKfak2SJGkYWGeIR8QFmXlkRPyGcq/01WTm0xurTJIkrdNALfG3tV5f1HQhkiRp/axzdHpm3tmavRdYmJm3ApsAewB3NFybJElah04vMbsS2LT1TPFLgdcCZzdVlCRJGlinIR6Z+TDwUuC/MvNwYLfmypIkSQPpOMQj4tmU68Mvbq3r9PI0SZLUgE5D/CTgfcB3M/PmiNgRuLy5siRJ0kA6vU78CuCKtuUFlBu/SJKkLhnoOvHPZOZJEfF9+r9O/NDGKpMkSes0UEu8517pn2y6EEmStH7WGeKZ2fOQkznA0sxcBRARYynXi0uSpC7pdGDbpcDmbcubAT8Z/HIkSVKnOg3xTTPzoZ6F1vzm69hekiQ1rNMQ/2tEPLNnISKeBSxtpiRJktSJTm/YchJwYUT03C99W+DlzZQkSZI60el14tdGxK7ALkAAv8vM5Y1WJkmS1qmj7vSI2Bx4D/C2zPwNMCMifDypJEld1Ok58a8By4Bnt5YXAR9ppCJJktSRTkP8SZn5cWA5QGYupXSrS5KkLuk0xJdFxGa0br0aEU8CHm2sKkmSNKBOR6d/EPgRMC0ivgXsCxzbVFGSJGlgA4Z4RATwO+ClwD6UbvS3Zea9DdcmSZLWYcAQz8yMiIsy81nAxUNQkyRJ6kCn58Svjoi9Gq1EkiStl07Pif8jcEJE3AL8ldKlnpn59KYKkyRJ69ZpiB/caBWSJGm9rTPEI2JT4ATgycBvgLMyc8VQFCZJktZtoHPiXwdmUgL8YOC0xiuSJEkdGag7fbfMfBpARJwF/LL5kiRJUicGaon/7UlldqNLkjS8DNQS3yMi/tKaD2Cz1nLP6PQtG61OkiSt1TpDPDPHDlUhkiRp/XR6sxdJkjTMGOKSJFXKEJckqVKGuCRJlTLEJUmqlCEuSVKlDHFJkipliEuSVClDXJKkShnikiRVyhCXJKlShrgkSZVqNMQj4qCImBcR8yPivevY7oiIyIiY2WQ9kiSNJI2FeESMBc4ADgZ2A46OiN362W4L4K3ANU3VIknSSNRkS3xvYH5mLsjMZcD5wGH9bPdh4OPAIw3WIknSiNNkiE8BFrYtL2qt+5uIeAYwLTN/sK4dRcTxETEnIubcc889g1+pJEkVajLEo591+bc3I8YAnwbeMdCOMvPMzJyZmTMnT548iCVKklSvJkN8ETCtbXkqcEfb8hbA7sBPI+IWYB9gloPbJEnqTJMhfi2wU0TsEBETgKOAWT1vZuaSzJyUmTMycwZwNXBoZs5psCZJkkaMxkI8M1cAJwKXAL8FLsjMmyPiQxFxaFOfK0nSaDGuyZ1n5mxgdp91H1jLtgc0WYskSSONd2yTJKlShrgkSZUyxCVJqpQhLklSpQxxSZIqZYhLklQpQ1ySpEoZ4pIkVcoQlySpUoa4JEmVMsQlSaqUIS5JUqUMcUmSKmWIS5JUKUNckqRKGeKSJFXKEJckqVKGuCRJlTLEJUmqlCEuSVKlDHFJkipliEuSVClDXJKkShnikiRVyhCXJKlShrgkSZUyxCVJqpQhLklSpQxxSZIqZYhLklQpQ1ySpEoZ4pIkVcoQlySpUoa4JEmVMsQlSaqUIS5JUqUMcUmSKmWIS5JUKUNckqRKGeKSJFXKEJckqVKGuCRJlTLEJUmqlCEuSVKlDHFJkipliEuSVClDXJKkShnikiRVyhCXJKlShrgkSZUyxCVJqpQhLklSpQxxSZIqZYhLklQpQ1ySpEoZ4pIkVcoQlySpUoa4JEmVMsQlSaqUIS5JUqUMcUmSKmWIS5JUKUNckqRKGeKSJFXKEJckqVKGuCRJlTLEJUmqlCEuSVKlDHFJkipliEuSVKlGQzwiDoqIeRExPyLe28/7b4+IuRFxY0RcGhHbN1mPJEkjSWMhHhFjgTOAg4HdgKMjYrc+m/0amJmZTwe+A3y8qXokSRppmmyJ7w3Mz8wFmbkMOB84rH2DzLw8Mx9uLV4NTG2wHkmSRpQmQ3wKsLBteVFr3docB/ywwXokSRpRxjW47+hnXfa7YcQxwExg/7W8fzxwPMD06dMHqz5JkqrWZEt8ETCtbXkqcEffjSLiecD7gUMz89H+dpSZZ2bmzMycOXny5EaKlSSpNk2G+LXAThGxQ0RMAI4CZrVvEBHPAL5ECfC7G6xFkqQRp7EQz8wVwInAJcBvgQsy8+aI+FBEHNra7BPAY4ELI+L6iJi1lt1JkqQ+mjwnTmbOBmb3WfeBtvnnNfn5kiSNZN6xTZKkShnikiRVyhCXJKlShrgkSZUyxCVJqpQhLklSpQxxSZIqZYhLklQpQ1ySpEoZ4pIkVcoQlySpUoa4JEmVMsQlSaqUIS5JUqUMcUmSKmWIS5JUKUNckqRKGeKSJFXKEJckqVKGuCRJlTLEJUmqlCEuSVKlDHFJkipliEuSVClDXJKkShnikiRVyhCXJKlShrgkSZUyxCVJqpQhLklSpQxxSZIqZYhLklQpQ1ySpEoZ4pIkVcoQlySpUoa4JEmVMsQlSaqUIS5JUqUMcUmSKmWIS5JUKUNckqRKGeKSJFXKEJckqVKGuCRJlTLEJUmqlCEuSVKlDHFJkipliEuSVClDXJKkShnikiRVyhCXJKlShrgkSZUyxCVJqpQhLklSpQxxSZIqZYhLklQpQ1ySpEoZ4pIkVcoQlySpUoa4JEmVMsQlSaqUIS5JUqUMcUmSKmWIS5JUKUNckqRKGeKSJFVqXLcLkCTV4fjjYfZsWLYMVq2C978fnvOcblc1uhnikqR1mjQJ9tsP/vxn+OMfYfx4+NWv4OlPN8S7zRCXJK3TJpvAFVesvm7MGPjKV2D6dFi5skwHHwy77tqdGkeryMxu17BeZs6cmXPmzOl2GZI0qu24I/zpT6uve9zj4OyzS1f7qlWwxx6w005dKW9EiYjrMnNmf+812hKPiIOAzwJjga9k5kf7vL8J8A3gWfaH0L0AAAgfSURBVMB9wMsz85Yma5IkbbybboIlS2Ds2DL90z/BjTfC4Yf3brP33nDNNd2rcTRoLMQjYixwBvB8YBFwbUTMysy5bZsdB9yfmU+OiKOAjwEvb6omSdLg2HzzMvW4/PLSMh8zpkxvfztcdlkJ9Z4O376vDzwAJ5xQts+EceNg//17l3usa36zzcovERG9n913PqKZYzAcNNkS3xuYn5kLACLifOAwoD3EDwNObc1/B/hcRETW1scvSaPcE55Qph5HHQX33AMLFpTlniDteb3++vJ61VVDU19E7y8AkyaVwXl9Ax/gkUfKGIBp09asudPXLbaA73632e/To8kQnwIsbFteBPzd2rbJzBURsQTYCri3wbokSQ17/evLtDaZJeCXLSvht2IF/PjHJVx7tLeg+5tftKj84pBZzsH3vK5t/s47y+dsskn/7z/6KNxwA2y33dp7DzJ7p7W9P3bshh2zDdFkiPfXgdG3hd3JNkTE8cDxANOnT9/4yiRJXRUBT3rS6ut23707tdSsyTu2LQKmtS1PBe5Y2zYRMQ54HLC4744y88zMnJmZMydPntxQuZIk1aXJEL8W2CkidoiICcBRwKw+28wCXtOaPwK4zPPhkiR1prHu9NY57hOBSyiXmH01M2+OiA8BczJzFnAWcE5EzKe0wI9qqh5JkkaaRq8Tz8zZwOw+6z7QNv8I8M9N1iBJ0kjlU8wkSaqUIS5JUqUMcUmSKmWIS5JUKUNckqRKGeKSJFXKEJckqVKGuCRJlTLEJUmqlCEuSVKlDHFJkipliEuSVClDXJKkShnikiRVKjKz2zWsl4i4B7h1EHc5Cbh3EPc3WnkcN57HcON5DDeex3DjDfYx3D4zJ/f3RnUhPtgiYk5mzux2HbXzOG48j+HG8xhuPI/hxhvKY2h3uiRJlTLEJUmqlCEOZ3a7gBHC47jxPIYbz2O48TyGG2/IjuGoPycuSVKtbIlLklSpURPiEXFQRMyLiPkR8d5+3t8kIr7dev+aiJgx9FUObx0cw7dHxNyIuDEiLo2I7btR53A20DFs2+6IiMiIcJRwPzo5jhFxZOvv480Rce5Q1zjcdfDveXpEXB4Rv279mz6kG3UOVxHx1Yi4OyJuWsv7ERGnt47vjRHxzEYKycwRPwFjgT8COwITgBuA3fps8ybgi635o4Bvd7vu4TR1eAz/Edi8Nf9Gj+H6H8PWdlsAVwJXAzO7Xfdwmzr8u7gT8Gvg8a3lrbtd93CaOjyGZwJvbM3vBtzS7bqH0wTsBzwTuGkt7x8C/BAIYB/gmibqGC0t8b2B+Zm5IDOXAecDh/XZ5jDg66357wDPjYgYwhqHuwGPYWZenpkPtxavBqYOcY3DXSd/DwE+DHwceGQoi6tIJ8fx9cAZmXk/QGbePcQ1DnedHMMEtmzNPw64YwjrG/Yy80pg8To2OQz4RhZXAxMjYtvBrmO0hPgUYGHb8qLWun63ycwVwBJgqyGprg6dHMN2x1F+C1WvAY9hRDwDmJaZPxjKwirTyd/FnYGdI+L/IuLqiDhoyKqrQyfH8FTgmIhYBMwG3jI0pY0Y6/t/5gYZN9g7HKb6a1H3HZbfyTajWcfHJyKOAWYC+zdaUX3WeQwjYgzwaeDYoSqoUp38XRxH6VI/gNIjdFVE7J6ZDzRcWy06OYZHA2dn5mkR8WzgnNYxXNV8eSPCkGTKaGmJLwKmtS1PZc2uob9tExHjKN1H6+oqGW06OYZExPOA9wOHZuajQ1RbLQY6hlsAuwM/jYhbKOfRZjm4bQ2d/nv+XmYuz8w/AfMooa6ik2N4HHABQGb+AtiUck9wdaaj/zM31mgJ8WuBnSJih4iYQBm4NqvPNrOA17TmjwAuy9boBAEdHMNWV/CXKAHuOcg1rfMYZuaSzJyUmTMycwZlXMGhmTmnO+UOW538e76IMtCSiJhE6V5fMKRVDm+dHMPbgOcCRMRTKCF+z5BWWbdZwKtbo9T3AZZk5p2D/SGjojs9M1dExInAJZRRmV/NzJsj4kPAnMycBZxF6S6aT2mBH9W9ioefDo/hJ4DHAhe2xgTelpmHdq3oYabDY6gBdHgcLwEOjIi5wErgXZl5X/eqHl46PIbvAL4cEf+P0g18rA2bXhFxHuV0zaTWuIEPAuMBMvOLlHEEhwDzgYeB1zZSh38mkiTVabR0p0uSNOIY4pIkVcoQlySpUoa4JEmVMsQlSaqUIS6NMhGxMiKuj4ibIuL7ETFxkPd/bER8rjV/akS8czD3L6mXIS6NPkszc8/M3J1yT4Q3d7sgSRvGEJdGt1/Q9lCGiHhXRFzbev7xv7Wtf3Vr3Q0RcU5r3Ysj4prW86Z/EhFP7EL90qg2Ku7YJmlNETGWclvNs1rLB1LuL7435eENsyJiP+A+yv3w983MeyPiCa1d/AzYJzMzIl4HvJtyly9JQ8QQl0afzSLiemAGcB3w49b6A1vTr1vLj6WE+h7AdzLzXoDM7Hkw0FTg261nJE8A/jQk1Uv6G7vTpdFnaWbuCWxPCd+ec+IB/GfrfPmemfnkzDyrtb6/+zP/F/C5zHwa8AbKAzIkDSFDXBqlMnMJ8FbgnRExnvIwjH+JiMcCRMSUiNgauBQ4MiK2aq3v6U5/HHB7a/41SBpydqdLo1hm/joibgCOysxzWo+c/EXrKXQPAce0nm7178AVEbGS0t1+LHAq5Yl1t1Mem7pDN76DNJr5FDNJkipld7okSZUyxCVJqpQhLklSpQxxSZIqZYhLklQpQ1ySpEoZ4pIkVcoQlySpUv8fMHlmQGSv4eMAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 576x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAfkAAAHwCAYAAACluRYsAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjMsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+AADFEAAAgAElEQVR4nOzdd3hUZf7+8feH3rsVUFFQBAsiglipgoiCqIiIChbUVVFxd+0o2HbR3/J1ld0VWURFwIqAgqDUtQCigkoTEJEAKiUUxdDy/P54BgwhZZLMyZlyv65rrsyZOTnzmUngzlPOc8w5h4iIiCSfEmEXICIiIsFQyIuIiCQphbyIiEiSUsiLiIgkKYW8iIhIklLIi4iIJCmFvEiEmV1tZlPDriOemNmvZnZsCK97jJk5MytV3K8dBDNbZGatCvF9+p2UIlHIS1wysx/M7PdIyPxkZiPNrFKQr+mce805d0GQr5GVmZ1lZtPNbLuZbTWziWbWqLheP4d6ZprZjVkfc85Vcs59H9DrHW9mb5rZxsj7/9rM+ptZySBer7Aif2zUL8oxnHONnXMz83mdg/6wKe7fSUk+CnmJZxc75yoBTYDTgPtDrqdQcmqNmllLYCowHjgSqAcsBD4JouUcby1iMzsOmAusAU52zlUFrgCaAZVj/Fqhvfd4+9wl9SjkJe45534CpuDDHgAzK2tmz5jZj2b2s5n9x8zKZ3m+i5ktMLNtZrbSzDpGHq9qZv81s/VmttbMHt/XcjSz3mb2ceT+f8zsmax1mNl4M+sfuX+kmb1tZhvMbJWZ9cuy36Nm9paZjTKzbUDvHN7WYOAV59yzzrntzrnNzrmHgDnAo5HjtDKzNDN7INLa/cHMro7mM8jyvfea2U/AS2ZW3czei9ScHrlfJ7L/E8C5wPOR3pPnI4/vb8VGelOGmtn7kd6HuZGw3lfPBWa2LNIq/5eZzcreM5DFQOBT51x/59z6yM95mXOup3NuS5b9ro68v41m9mCW12puZp+Z2ZbIz/J5MyuT5XlnZreZ2XJgeeSxZ81sTeR34gszOzfL/iUjn/PKyHv7wszqmtnsyC4LI5/LlZH9O0d+v7aY2admdkqWY/0Q+dy/Bn4zs1KRx9plqX1+pI6fzewfkW/d91pbIq/VMuvvZOR7G5vZh2a2OfK9D+Ty+Yp4zjnddIu7G/AD0C5yvw7wDfBsluf/D5gA1MC3/CYCT0Weaw5sBdrj/5CtDTSMPPcu8AJQETgUmAfcHHmuN/Bx5P55+FamRbarA7/jW90lgC+AAUAZ4Fjge6BDZN9Hgd1A18i+5bO9twrAXqB1Du+7D7A+cr8VsAf4B1AWOB/4DTghis9g3/f+PfK95YGawGWR168MvAm8m+W1ZwI3ZqvHAfUj90cCmyOfbyngNWBs5LlawDagW+S5OyOfwY25/Hx/Avrk8fM/JvLaL0ZqPxXYCZwYef504MzIax0DLAHuylb3h5HPpnzksV6Rz6AUcE+khnKR5/6C/x07AbDI69XM/hlEtpsCvwAtgJLAdfjf17JZfncXAHWzvPYP/PH7/BlwTeR+JeDMbO+5VJbX6s0fv5OVgfWR2stFtluE/W9Vt/i+hV6AbrrldIv8p/grsD3yH980oFrkOcOH3XFZ9m8JrIrcfwEYksMxD4sERfksj10FzIjcz/ofqgE/AudFtm8CpkfutwB+zHbs+4GXIvcfBWbn8d7qRN5Twxye6wjsjtxvhQ/qilmefwN4OIrPoBWwa1+I5VJHEyA9y/ZM8g/54Vme6wQsjdy/Fvgsy3OG/yMpt5DfDXTMo7Z9gVcny2PzgB657H8XMC5b3W3y+R1LB06N3F8GdMllv+wh/2/gsWz7LAPOz/K7e30Ov8/7Qn42viejVi7vObeQvwr4Ksh/d7ol303jRRLPujrnPjKz84HR+NbiFuAQfGv0CzPbt6/hW1XgW1CTcjje0UBpYH2W7yuBD6MDOOecmY3F/8c6G+gJjMpynCPNLGu3ckngf1m2DzpmFulAJnAEsDTbc0cAG7Pu65z7Lcv2anxvQn6fAcAG51zG/ifNKgBD8H9IVI88XNnMSjrn9uZRb1Y/Zbm/A98SJVLT/vcc+fzS8jjOJvx7LdTrmdnx+B6OZvjPoRS+dyWrA34GZnYPcGOkVgdUwf9Ogf+dWRlFPeB//teZ2R1ZHisTOW6Or53NDcAgYKmZrQIGOufei+J1C1KjCKAxeUkAzrlZ+FbkvjHyjfiu88bOuWqRW1XnJ+mB/w/2uIOPxBp8S75Wlu+r4pxrnMtLjwEuN7Oj8a33t7McZ1WWY1RzzlV2znXKWnYe7+c3fJftFTk83R3fa7FPdTOrmGX7KGBdFJ9BTjXcg++ObuGcq4IfkgD/x0GeNUdhPb6Hwh/Q/+VRJ/fd+Qg/dFBY/8b/gdQg8l4e4I/3sc/+9xMZf78X//lWd85Vww/p7Pue3H5ncrIGeCLbz7+Cc25MTq+dnXNuuXPuKvxw0d+BtyI/4/w+/4LUKAIo5CVx/B/Q3syaOOcy8WO1Q8zsUAAzq21mHSL7/hfoY2ZtzaxE5LmGzk/wmgr8PzOrEnnuuEhPwUGcc18BG4DhwBT3x4SwecC2yOSq8pFJWyeZ2RkFeD/34VuD/cyscmRS3OP4LveB2fYdaGZlIkHVGXgzis8gJ5XxfxhsMbMawCPZnv8ZP7+gMN4HTjazruZnlN8GHJ7H/o8AZ5nZ02Z2eKT++uYnK1aL4vUq4+cA/GpmDYFbo9h/D/7nWcrMBuBb8vsMBx4zswbmnWJmNSPPZf9cXgRuMbMWkX0rmtlFZhbVWQFm1svMDon8DPf9Tu2N1JZJ7j+D94DDzewu85MuK5tZi2heU1KXQl4SgnNuA/AKfjwafKtsBTDH/Az2j/CtVJxz8/AT2IbgW2uz8F2s4MeOywCL8d3mb5F3t/EYoB1+uGBfLXuBi/Fj2qvwrerhQNUCvJ+PgQ74iWrr8d3wpwHnOOeWZ9n1p0id6/AT3W5xzu3r4s/1M8jF/+EnsW3Ez+L/INvzz+J7LtLN7J/RvpfI+9mI75kYjO+KbwTMx/ec5LT/SvwfNMcAi8xsK76nZD5+HkZ+/owfQtmOD93X89l/CjAZ+A7/WWdwYJf6P/DzHabi/3j4L/6zAj/H4uXITPruzrn5+Dkaz+N/NivI+QyK3HTEv+df8Z95D+dchnNuB/AE/jTKLWZ2ZtZvcs5tx08mvRj/e7EcaF2A15UUtG/msIjEGfMrpI1yzuXV7R2XzKwEkAZc7ZybEXY9IqlKLXkRiQkz62Bm1cysLH+Mkc8JuSyRlKaQF5FYaYmf/b0R36Xc1Tn3e7gliaS2wLrrzWwEfpLQL865k3J43vDjUZ3wp8b0ds59GUgxIiIiKSjIlvxI/AST3FwINIjc+uJPiREREZEYCSzknXOz8Utg5qYLfu1u55ybA1Qzs2gWxxAREZEohLniXW0OPIUlLfLY+uw7mllffGufihUrnt6wYcNiKVBERCQsq37ZTq2f0/hu146NzrlDCnOMMEM+++pUkMuKT865YcAwgGbNmrn58+cHWZeIiEjoprS6jPZpyynp13YolDBn16fh12Lepw5+wQ8REZGU92bnG/hH3yeLdIwwQ34CcG1kWcgzga2RZUdFRERS09at8OCDsGsX2ytX5/PTclx1O2qBddeb2Rj85S5rRa5G9Qj+CmA45/6Dv0pYJ/ySkDvwy5CKiIikps2boUMHWLiQqcecztxVZWlRr0aRDhlYyEeuspTX8w5/EQsREZHUtnEjtG8PixfDO+/w37U1gc10aVKbN4pwWK14JyIiEqaff4ZWrWDpUpgwATp3BqBFvRr0bHFUkQ4d5ux6ERERWbcO0tPh/fehTZuYHlohLyIiUkxGz/2R8QvWAlAu4zcyylUEoPT9r7F7eVlY/hkAi9dvo9ERVYr8euquFxERKSbjF6xl8fptHLJxPYMfv45O014HYHfpsgfs1+iIKnRpUrvIr6eWvIiISDFqXWob/3zhHtj7O9f9tRfXNWsW2Gsp5EVERIrJET+tZsCQO6CUg+nToUmTQF9PIS8iIlIctm9nwJA7KJG5F2bOhpMOugp7zCnkRUQk5WWdEBek41r35vcmTRlSDAEPmngnIiKyf0JcEOqtXspJS/2F1VZ27MYZHc8O5HVyopa8iIgIfkb76ze3jO1B58yBe++G2rXhma+hZMnYHj8fasmLiIgE4eOP4YILoFYtmDy52AMe1JIXEZEYKK4x7aDEavGZ/WbO9MvT1q7tZ9HXLvo574WhlryIiBRZkGPaxSFWi8/s9+abcPTRMGtWaAEPasmLiEiMBDKmnWh274bSpeGf/4Rt26B69VDLUUteREQkFsaPh5NPhrQ0P/4ecsCDQl5ERKTo3nwTLr8cqlWDSpXCrmY/ddeLiCSZMCbBxXziWiIZPRquuQZatoRJk6BK/HwOasmLiCSZMCbBxXziWqIYPx569YLzzoMPPoirgAe15EVEkpImwRWTc8+F22+Hv/0NKlQIu5qDqCUvIiJSUO++CxkZUKOGn0kfhwEPasmLiBS7oMfMU3p8vDj84x9wzz3w1FNw331hV5MnteRFRIpZ0GPmKTs+XhyeesoH/BVX+K9xTi15EZEQaMw8wTgHgwbBo49Cz57w8stQKv4jVC15ERGR/Pz8sx97790bXnklIQIe1JIXEYmpaMbbNWaeQJzzXw8/HObP9+vRl0ic9nHiVCoikgCiGW/XmHmCcA7uvBMee8xv16uXUAEPasmLiMScxtuTQGYm3HorDBsG/fv7wDcLu6oCS6w/SURERIK2dy/ccIMP+Pvvh2eeSciAB4W8iIjIga6/HkaO9DPpn3giYQMe1F0vIilGC9FIvtq2hYYNfSs+wSnkRSSl7JsYF1QQa1Jdgtq5ExYsgBYt4Nprw64mZhTyIpJyNDFODpCR4a8FP20aLF8OdeqEXVHMKORFRCR17dgBl14KU6fCCy8kVcCDQl5EkogWopEC+e03uPhimDkTRoyAPn3CrijmNLteRJKGFqKRAhk+HGbNgldfTcqAB7XkRSTJaLxdonbHHdCyJTRvHnYlgVFLXkREUsfmzX4MftUqv0RtEgc8KORFRCRVbNgArVvD5Ml+Fn0KUMiLSFIYPfdH5q7aHHYZEq9++glatYLvvoMJE+CCC8KuqFhoTF5EksK+WfWaVCcHWb/et+DXrIFJk/z9FKGWvIgkjRb1atCzxVFhlyHxpmJFOOoomDIlpQIe1JIXEZFk9eOPUKsWVKniAz6BLzRTWAp5EUlI2Re+0SI3coAVK6BNGzjzTHjjjZQMeFB3vYgkqOwL32iRG9lv6VI47zz4/Xd48MGwqwmVWvIikrC08I0c5Ntv/aVizWDGDDjppLArCpVCXkREkkNmJvToASVLwvTp/prwKU4hLyIiyaFECRg7FsqVg/r1w64mLijkRZJYNFdlS1SaaCf7zZkDH3wAjzyS8t3z2WninUgSi+aqbIlKE+0EgP/9D9q3h9deg61bw64m7qglL5LkNDlNktb06f568HXr+vvVqoVdUdxRS15ERBLP1Klw0UVQr56/JvyRR4ZdUVxSS14kSok4vq1xa0laW7ZA48b+inKHHBJ2NXFLLXmRKCXi+LbGrSXp/Pyz/9q9O8ydq4DPh1ryIgWg8W2REL3xBvTuDe+/7y80U7Jk2BXFPbXkRUQk/o0aBVddBc2a+ZtERSEvko/Rc3/kyhc+S7iuepGkMWIEXHstnH++H4OvXDnsihKGQl4kH/vG4jW+LRKCTz+FG27w58K/956/NrxETWPyIlHQWLxISFq2hGHD4Jpr/HK1UiBqyYuISPz517/gu+/81eRuukkBX0gKeRERiS9PPgm33QbPPRd2JQlP3fUpIBEXcYknWlBGpJg4BwMH+luvXjBkSNgVJTy15FNAIi7iEk804U6kGDgHDzzgA75PHxg5EkqpHVpU+gRThCaOiUhc27ULPvkEbr7Zj8eXUBs0FhTyIiISnsxMyMiAChX8NeHLl/eT7SQmFPJxKNZj6BpTFpG4lJkJt9wCS5fChx/6oJeYUn9IHIr1GLrGlEUk7uzdC9dfDy++COedB2XKhF1RUlJLPk5pDF1EktaePXDddTB6NAwaBA8/HHZFSUshLyIixeuuu3zAP/UU3Hdf2NUkNYW8iIgUrzvugMaN4dZbw64k6SnkY0AT5URE8pGRAa+84peoPeEEf5PAaeJdDGiinIhIHnbsgIsv9jPp584Nu5qUopZ8jGiinIhIDn791Qf87Nnw0ktw5plhV5RSFPIiIhKMbdugUyeYMwdefRV69gy7opSTcCG/+bddXPnCZ2GXcQCNoYuI5GD+fPjySxg7Fi6/POxqUlLChfyWHbvjLlQ1hi4iksXevVCyJLRpA6tWwWGHhV1Rykq4kAeNf4uIxK1ffoELL/Tnv19xhQI+ZAkZ8iIiEofWr4d27XzrvXr1sKsRFPIiIhILa9f67vm1a2HSJGjVKuyKhIDPkzezjma2zMxWmNlBaxea2VFmNsPMvjKzr82sU5D1iIhIANLT4fzzfUt+yhQFfBwJLOTNrCQwFLgQaARcZWaNsu32EPCGc+40oAfwr6DqERGRgFSrBr16+cvFnn122NVIFkF21zcHVjjnvgcws7FAF2Bxln0csG+afFVgXYD1iIhILC1fDjt3wkknwaOPhl2N5CDIkK8NrMmynQa0yLbPo8BUM7sDqAi0y+lAZtYX6AtQ6YjjYl6oiIgU0JIlfgy+Vi1YuBBKaJX0eBTkT8VyeMxl274KGOmcqwN0Al41s4Nqcs4Nc841c841K126dACliohI1L75xo/BOwevv66Aj2NB/mTSgLpZtutwcHf8DcAbAM65z4ByQK0AaxIRkaL46ito3RpKl4ZZs6BR9qlWEk+CDPnPgQZmVs/MyuAn1k3Its+PQFsAMzsRH/IbAqxJRESK4sknoWJFf8EZXS427gU2Ju+c22NmtwNTgJLACOfcIjMbBMx3zk0A7gFeNLO78V35vZ1z2bv0RUQkbM6BGYwcCZs3Q926+X6LhC/QxXCcc5OASdkeG5Dl/mJA51uIiMSz2bPhscfg7behShXfkpeEoNkSIiKSu2nToGNHv5Ldjh1hVyMFpJAXEZGcTZkCnTtD/fowcyYcfnjYFUkBKeRFRORgU6bAJZdAw4YwfTocemjYFUkhKORFRORgDRpAp06+u76WzmxOVAp5ERH5w7x5kJkJxx4L48ZBjRphVyRFoJAXERHv1VehZUt47rmwK5EYUciLiAiMGAHXXecvE3vjjWFXIzGikBcRSXX//jfccANccAG8957Og08iCnkRkVS2Zg3cfbc/Ve7dd6F8+bArkhgKdMU7ERGJc3Xr+gvNnHYalCkTdjUSY2rJi4ikoief9BPtAFq0UMAnKYW8iEgqcQ4GDIAHH/SL3EhSU3e9iEiqcA7uuw8GD4brr4dhw8KuSAKmlryISCpwDvr39wF/663w4otQsmTYVUnAFPIiIqnADKpWhTvvhKFDoYT++08F6q4XEUlmmZmwejXUqwePPOIfMwu3Jik2+lNORCRZ7d0LffpAs2bw008+3BXwKUUhLyKSjHbvhl694JVX/GI3uhZ8SlJ3vYhIstm1C666Ct55B/7+d/jrX8OuSEKikBcRSTbPPOMDfsgQuOuusKuRECnkRUSSTf/+0LgxdOkSdiUSMo3Ji4gkg99+g379ID0dypVTwAugkBcRSXzbt0OnTv78948/DrsaiSPqrhcRSWRbt8KFF8K8efDaa3DxxWFXJHFEIS8ikqjS06FDB/jqK3j9dbjssrArkjijkBcRSVS//grbtvmZ9GrBSw4U8iIiiSY9HapUgbp14ZtvoHTpsCuSOKWJdyIiiWT9ejj7bH+hGVDAS57UkhcRSRRpadCmDaxbB927h12NJACFvIhIIli92gf8xo0wdSqcdVbYFUkCUMiLiMS7PXv8aXKbN8OHH0Lz5mFXJAlCIS8iEu9KlYJ//hNq1ICmTcOuRhKIJt6JiMSrxYv9pWIB2rVTwEuBqSUvIhKPvv7aB3vp0nDppVC5ctgVSQJSS15EJN58+SW0bg1lysCMGQp4KTSFvIhIPJk3D9q2hUqVYNYsOP74sCuSBKaQFxGJJ5984ifYzZ4Nxx0XdjWS4BTyIiLxICPDf737bliwAI4+Otx6JCko5EVEwvbRR77V/tVXfltj8BIjCnkRkTBNngydO0PNmlC7dtjVSJJRyIuIhGXCBOjaFRo18rPoDz007IokySjkRUTC8MkncNll0KQJTJvmW/IiMaaQFxEJwxlnwL33+rXoq1cPuxpJUgp5EZHi9M47sGGDX+jm8cehSpWwK5IkppAXESkuw4fD5ZfDoEFhVyIpQiEvIlIchg6Fm26CDh1g8OCwq5EUoZAXEQnakCFw++1wySXw7rtQvnzYFUmKUMiLiARpxw74z3/8TPo334SyZcOuSFKILjUrIhKUzEyoUAH+9z+/Hn0p/ZcrxUsteRGRWHMOHnoIrrkG9u71i9wo4CUECnkRkVhyzp///sQTULEimIVdkaQwhbyISKw4568i9/TTcNttfiy+hP6blfDot09EJFb+8hd49lkf9M89p4CX0GmQSEQkVi65xHfRP/qouuklLujPTBGRotizB6ZM8ffPOw8GDlTAS9xQyIuIFNbu3dCrF3TsCF99FXY1IgdRd72ISGHs2gU9esC4cX6i3WmnhV2RyEEU8iIiBbVzp7/QzHvv+Yl2/fqFXZFIjhTyIiIFNWkSvP8+/PvfcMstYVcjkiuFvIhIQV16KXz9NZx0UtiViORJE+9ERKKxfTtcdBF8+qnfVsBLAlDIi4jkZ+tWfx34KVNg7dqwqxGJmrrrRUTysnmzD/iFC/2lYi+9NOyKRKKmkBcRyU16OrRtC4sXwzvvQOfOYVckUiDqrhcRyU3lynDyyTBhggJeEpJa8iIi2a1b55emPeIIeOWVsKsRKTSFvIhIVmvWQJs2UKMGzJmjdegloSnkRUT2+eEHH/CbNvkWvAJeEpxCXkQEYMUKH/Dbt8NHH8EZZ4RdkUiRKeRFRABuuw127IAZM6BJk7CrEYkJhbyICMDLL/tu+saNw65EJGZ0Cp2IpK6FC6FvX39d+MMPV8BL0lFLXkRS0xdfQPv2ULEi/PQT1K0bdkUiMaeWvIiknjlz/Ep2VarA7NkKeElaCnkRSS0ff+xb8LVq+YCvVy/sikQCo5AXkdRSqhQ0bAizZsFRR4VdjUigFPIikhpWr/ZfzzwT5s2D2rXDrUekGAQa8mbW0cyWmdkKM7svl326m9liM1tkZqODrEdEUtSkSXDCCX+sQ6+V7CRFBDa73sxKAkOB9kAa8LmZTXDOLc6yTwPgfuBs51y6mR0aVD0ikqLGj4crrvBXk7voorCrESlWQbbkmwMrnHPfO+d2AWOBLtn2uQkY6pxLB3DO/RJgPSKSat58Ey6/HJo2hWnToGbNsCsSKVZRhbyZlTGz+gU8dm1gTZbttMhjWR0PHG9mn5jZHDPrWMDXEBHJ2fffQ8+e0KIFTJ0K1aqFXZFIscs35M3sIuAb4MPIdhMzGxfFsXMa9HLZtksBDYBWwFXAcDM76F+imfU1s/lmNn/37t1RvLSIpLxjj4XXX4cPPvDnw4ukoGha8oOAFsAWAOfcAiCaVn0akHWFiTrAuhz2Ge+c2+2cWwUsw4f+AZxzw5xzzZxzzUqXLh3FS4tIyho+HGbO9Pe7dYNKlUItRyRM0YT8bufclmyPZW+R5+RzoIGZ1TOzMkAPYEK2fd4FWgOYWS189/33URxbRORgzz8PN90E//532JWIxIVoQn6JmXUHSkQC+/+AOfl9k3NuD3A7MAVYArzhnFtkZoPM7JLIblOATWa2GJgB/MU5t6lQ70REUts//gF33AFduvxxqpxIiovmFLrbgQFAJvAOPpjvj+bgzrlJwKRsjw3Ict8B/SM3EZHCeeopeOABf6rca6+BhvVEgOhCvoNz7l7g3n0PmFk3fOCLiITLOViyBK6+GkaO9MvWiggQXXf9Qzk89mCsCxERKRDnYMsWv3rdiBHw8ssKeJFscv0XYWYdgI5AbTP7R5anquC77kVEwuEc/OUv8O67MHeuFrkRyUVef/b+AnwLZACLsjy+HchxHXoRkcA5B3feCc89B7ffDjVqhF2RSNzKNeSdc18BX5nZa865jGKsSUQkZ5mZ8Kc/wQsvQP/+8MwzutiMSB6iGcCqbWZPAI2AcvsedM4dH1hVIiI5efJJH/D33w9PPKGAF8lHNCE/EngceAa4EOiDxuRFJAy33OK752+9VQEvEoVoZtdXcM5NAXDOrXTOPURklToRkcDt3g1PPw07d0KtWr67XgEvEpVoWvI7zcyAlWZ2C7AW0HXfRSR4O3dCjx5+Fn2DBtC1a9gViSSUaEL+bqAS0A94AqgKXB9kUSIiZGTAZZfBpEl+Jr0CXqTA8g1559zcyN3twDUAZlYnyKJEJMXt2OFD/cMP/US7vn3DrkgkIeU5Jm9mZ5hZ18gV4jCzxmb2ClFcoEZEpNB++AG++MKvZKeAFym0vFa8ewq4DFgIPGRm44A7gb8DtxRPeSKSUnbuhLJloVEjWLkSqlULuyKRhJZXd30X4FTn3O9mVgNYF9leVjyliUhK2bIFOnb03fT33aeAF4mBvLrrM5xzvwM45zYDSxXwIhKIzZuhXTv48kto2DDsakSSRl4t+WPNbN/lZA04Jss2zrlugVYmIqlhwwYf8MuWwbhxcNFFYVckkjTyCvnLsm0/H2QhIpKCdu2Ctm1h+XKYMAEuuCDsikSSSl4XqJlWnIWISAoqUwbuvhuOOQZaayFNkViLZjEcEZHY+vFHWLEC2rSBPn3CrkYkaSnkRaR4rVrlwz0jA77/HsqXD7sikaQVdcibWVnn3M4gixGRJLev9f7rr341OwW8SKDyvQqdmTU3s2+A5ZHtU83sucArE5HksnQpnHce/P47zJgBp58edkUiSS+aS83+E+gMbAJwzi1El5oVkYJ66SXIzISZM+HUU8OuRiQlRBPyJZxzq7M9tjeIYkQkCTnnvz71FMyfD40bh1uPSAqJJuTXmJAtNZIAACAASURBVFlzwJlZSTO7C/gu4LpEJBnMn++75VevhhIloI4uYClSnKIJ+VuB/sBRwM/AmZHHRERyN2eOX+gmPf2P1ryIFKtoZtfvcc71CLwSEUke//sfdOoEhx8O06dD3bphVySSkqJpyX9uZpPM7Dozqxx4RSKS2ObM8VeTq1MHZs1SwIuEKN+Qd84dBzwOnA58Y2bvmpla9iKSs4YNoVs3P4v+yCPDrkYkpUXTksc596lzrh/QFNgGvBZoVSKSeD75xJ8DX60avPoqHHZY2BWJpLxoFsOpZGZXm9lEYB6wATgr8MpEJHGMG+cvMPPQQ2FXIiJZRDPx7ltgIjDYOfe/gOsRkUTzxhvQsyeccQY8/HDY1YhIFtGE/LHOuczAKxGRxDNqFFx3HZx1FkyaBJU1N1cknuQa8mb2/5xz9wBvm9lBJ7k657oFWpmIxLft2+Gee+D882HiRKhYMeyKRCSbvFryr0e+Pl8chYhIgqlc2Z8id9RRUKFC2NWISA5yDXnn3LzI3ROdcwcEvZndDkwLsjARiVPPPQcbNsCgQf50ORGJW9GcQnd9Do/dEOtCRCQBPPMM9OsH33wDe3WdKpF4l9eY/JVAD6Cemb2T5anKwJagCxOROPPkk/Dgg9C9u59wV7Jk2BWJSD7yGpOfh7+GfB1gaJbHtwNfBVmUiMSZQYPgkUegVy9/XfhS0ZyYIyJhy2tMfhWwCvio+MoRkbh03HFw443wn/+oBS+SQHIdkzezWZGv6Wa2Ocst3cw2F1+JIhIK5/zYO8DVV8OLLyrgRRJMXhPvWke+1gIOyXLbty0iySozE+64A04/HRYtCrsaESmkXEM+yyp3dYGSzrm9QEvgZkCrXogkq8xMuPlmGDoU7roLGjUKuyIRKaRoTqF7F3BmdhzwCnAiMDrQqkQkHHv3wvXXw/Dhfib93/8OZmFXJSKFFE3IZzrndgPdgP9zzt0B1A62LBEJxejR8PLLfjb9448r4EUSXDTnwewxsyuAa4CukcdKB1eSiISmVy845BDo2DHsSkQkBqJd8a41/lKz35tZPWBMsGWJSLHZudOPwa9c6VvuCniRpJFvyDvnvgX6AfPNrCGwxjn3ROCViUjwMjKgWzcYNgz+97+wqxGRGMu3u97MzgVeBdYCBhxuZtc45z4JujgRCdCOHdClC0yb5kO+d++wKxKRGItmTH4I0Mk5txjAzE7Eh36zIAsTkQD9+itcfLG/VOxLL8F114VdkYgEIJqQL7Mv4AGcc0vMrEyANYlI0Jzzp8uNGgU9e4ZdjYgEJJqQ/9LMXsC33gGuRheoEUlMW7ZA6dJQuTLMnAklopl7KyKJKpp/4bcAK4G/AvcC3+NXvRORRLJpE7RtC5df7lvyCniRpJdnS97MTgaOA8Y55wYXT0kiEnO//ALt28OyZfDYY1rkRiRF5HUVugfwS9peDXxoZtcXW1UiEjvr10Pr1rB8OUycCJ06hV2RiBSTvFryVwOnOOd+M7NDgEnAiOIpS0Ri5qqrYPVqmDQJWrUKuxoRKUZ5hfxO59xvAM65DWamATyRRPSvf0F6Opx9dtiViEgxyyvkjzWzdyL3DTguyzbOuW6BViYihff99/D663DffbpUrEgKyyvkL8u2/XyQhYhIjCxfDm3a+BXtrrkG6tQJuyIRCUmuIe+cm1achYhIDCxZ4gN+zx6YPl0BL5LiNM4ukiy++QbOP9+fAz9zJpx6atgViUjIFPIiyWLlSqhQwa9H37hx2NWISByIOuTNrGyQhYhIIW3b5r927QpLl8IJJ4Rbj4jEjXxD3syam9k3wPLI9qlm9lzglYlI/j79FI49Ft5/32+XKxduPSISV6Jpyf8T6AxsAnDOLQRaB1mUiERh9my44AKoUUPj7yKSo2hCvoRzbnW2x/YGUYyIRGnaNOjYEerW9WPwmkUvIjmIJuTXmFlzwJlZSTO7C/gu4LpEJDdLl0LnznDccX4W/RFHhF2RiMSpaEL+VqA/cBTwM3Bm5DERCcMJJ8ATT8CMGXDYYWFXIyJxLM9LzQI4534BehRDLSKSl/Hj4fjj4cQToX//sKsRkQSQb8ib2YuAy/64c65vIBWJyMHGjoVevaBLF3j77bCrEZEEkW/IAx9luV8OuBRYE0w5InKQV1+F3r3hnHNg5MiwqxGRBBJNd/3rWbfN7FXgw8AqEpE/jBgBN94IrVvDhAlQsWLYFYlIAinMsrb1gKNjXYiIZJOZ6VvxHTrAe+8p4EWkwKIZk0/njzH5EsBm4L4gixJJebt3Q+nSMHGi/1pWq0qLSMHlGfJmZsCpwNrIQ5nOuYMm4YlIDA0e7Lvmp0yBSpXCrkZEElie3fWRQB/nnNsbuSngRYL0+ONw771+JTu13kWkiKIZk59nZk0Dr0QklTkHAwbAww/DNdfAqFFQKpqTX0REcpfr/yJmVso5twc4B7jJzFYCvwGGb+Qr+EViZfBgeOwxuOEGeOEFKFky7IpEJAnk1VSYBzQFuhb24GbWEXgWKAkMd879LZf9LgfeBM5wzs0v7OuJJKxu3fx14R97DEoU5qQXEZGD5RXyBuCcW1mYA5tZSWAo0B5IAz43swnOucXZ9qsM9APmFuZ1RBJWZia89RZccQU0aODXoxcRiaG8Qv4QM8t1gWzn3D/yOXZzYIVz7nsAMxsLdAEWZ9vvMWAw8Of8yxVJEpmZcPPNMHw4lC8PF18cdkUikoTy6hcsCVQCKudyy09tDlz+Ni3y2H5mdhpQ1zn3XgFqFklse/dCnz4+4B96yF82VkQkAHm15Nc75wYV4diWw2P7T8EzsxLAEKB3vgcy6wv0Bah0xHFFKEkkZHv2+NnzY8fCoEF+Nr2ISEDyasnnFNIFkQbUzbJdB1iXZbsycBIw08x+wF+nfoKZNct+IOfcMOdcM+dcs9KlSxexLJEQffmlv4rc3/+ugBeRwOXVkm9bxGN/DjQws3r4FfN6AD33Pemc2wrU2rdtZjOBP2t2vSQl58AMmjeHpUvh2GPDrkhEUkCuLXnn3OaiHDhyjv3twBRgCfCGc26RmQ0ys0uKcmyRhPL7735i3dixflsBLyLFJNAltZxzk4BJ2R4bkMu+rYKsRSQUv/0GXbrA9OnQtdBLToiIFIrWzRQJyvbtfub8xx/DyJFw7bVhVyQiKUYhLxKEjAx/Hfh58+C116BHj7ArEpEUpPUzRYJQtiy0bQuvv66AF5HQqCUvEkubNsFPP0Hjxn4dehGRECnkRWLll1+gXTvYuhW++07XgxeR0CnkRWJh/XrfPf/DDzBxogJeROKCQl6kqNLSoE0bWLcOJk+G888PuyIREUAhL1J0AwfCzz/D1Klw1llhVyMisp9m14sU1bPP+nPhFfAiEmcU8iKFsWwZXHqpn2RXoQKcfHLYFYmIHETd9SIFtXixH4PPzPTj8FWrhl2RiEiO1JIXKYivv4ZWrfwV5WbOhBNPDLsiEZFcKeRForVgAbRuDWXKwKxZ0KhR2BWJiORJIS8SrWrV/Ep2s2fD8ceHXY2ISL4U8iL5WbbMj78fc4xvwet68CKSIBTyInmZNQtOPx0ef9xvm4Vbj4hIASjkRXLz0Udw4YVw9NHQt2/Y1YiIFJhCXiQnH3wAnTtD/fowYwYcfnjYFYmIFJhCXiS7LVvgyiv97PkZM+DQQ8OuSESkULQYjkh21arBhAlwyilQvXrY1YiIFJpCXmSfMWMgIwP69NGV5EQkKai7XgTg5ZehVy8YNcqfLicikgQU8iLDh/vWe5s2MHEilNA/CxFJDvrfTFLbv/4FN90EHTr4cfgKFcKuSEQkZhTykto2bYJLLoF334Xy5cOuRkQkpjTxTlLTzz/DYYfBww/D3r1QsmTYFYmIxJxa8pJ6HnvMXyJ25Uq/rYAXkSSlkJfU4Rw89BAMGAAXX+wvOCMiksTUXS+pwTm49154+mm48UZ44QXNoheRpKf/5SQ1jBjhA/5Pf1LAi0jKUEteUkOvXn6Rmxtv1OViRSRlqDkjyWvvXj/JbtMmKFvWnw+vgBeRFKKQl+S0Zw/07u0n2b31VtjViIiEQt31knx27/bd82+8AY8/DjffHHZFIiKhUMhLctm1C3r0gHHj/ES7P/857IpEREKjkJfksnkzfP01PPss9OsXdjUiIqFSyEtyyMiA0qXh8MN9yOtCMyIimngnSeC336BTJ+jb128r4EVEAIW8JLrt2+HCC2HWLGjdOuxqRETiirrrJXFt3eoDft48GD0arrwy7IpEROKKQl4Sk3PQtSvMnw9vvgmXXhp2RSIicUchL4nJzF8LfscO6Nw57GpEROKSQl4Sy88/w4wZ/lz4Nm3CrkZEJK4p5CVxrFsHbdvCmjV+kt1hh4VdkYhIXFPIS2JYs8a33H/6CSZPVsCLiERBIS/x74cffMBv2gRTp0LLlmFXJCKSEBTyEv+mTIEtW+Cjj+CMM8KuRkQkYWgxHIlfe/b4rzffDEuXKuBFRApIIS/xadEiaNzYL3QDcOih4dYjIpKA1F0v8WfhQmjXzl9wpkqVsKsREUlYaslLfPniC396XLlyfj36hg3DrkhEJGEp5CV+LFniz4OvUgVmz4YGDcKuSEQkoSnkJX7Urw/XXusDvl69sKsREUl4CnkJ3yef+OVqS5eGf/4Tjjoq7IpERJKCQl7C9eGH0L499OsXdiUiIklHIS/hmTQJLr7Yj70//3zY1YiIJB2FvIRj/Hh/PfjGjWH6dDjkkLArEhFJOgp5KX579vhrwTdtCtOmQc2aYVckIpKUtBiOFC/noFQpvx59xYpa7EZEJEBqyUvxGTkSevXyLfkjjlDAi4gETCEvxWPYMOjTBzZsgN27w65GRCQlKOQleM8/768k16kTTJgA5cuHXZGISEpQyEuwnnsO7rgDunSBd97xa9KLiEixUMhLsE47DXr3hjffhLJlw65GRCSlKOQl9pyDOXP8/XPOgZde8kvWiohIsVLIS2w5Bw89BC1bwkcfhV2NiEhK03nyEjvOwV/+Av/v/0HfvtCmTdgViYikNIW8xIZzcOedfqLd7bf7q8mZhV2ViEhKU3e9xMasWT7g+/dXwIuIxAm15CU2WrWCjz+Gs85SwIuIxAm15KXw9uzxi9x8/LHfPvtsBbyISBxRyEvh7N4NPXv65Wrnzg27GhERyYG666Xgdu6EHj3g3Xf9TPr+/cOuSEREcqCQl4LZuRMuuwzef/+PmfQiIhKXFPJSMKVKQdWq8MIL/lx4ERGJWwp5ic5vv8G2bf468KNGaYKdiEgC0MQ7yd+2bdCxI7Rr5yfcKeBFRBKCWvKSty1b4MIL4fPPYfRoXWhGRCSBBNqSN7OOZrbMzFaY2X05PN/fzBab2ddmNs3Mjg6yHimgzZt96/2LL+Ctt6B797ArEhGRAggs5M2sJDAUuBBoBFxlZo2y7fYV0Mw5dwrwFjA4qHqkEO64A779FsaNg65dw65GREQKKMiWfHNghXPue+fcLmAs0CXrDs65Gc65HZHNOUCdAOuRghoyBD74AC66KOxKRESkEIIM+drAmizbaZHHcnMDMDnAeiQaa9fCXXf5CXaHHurXpBcRkYQUZMjnNAXb5bijWS+gGfB0Ls/3NbP5ZjZ/9+7dMSxRDvDjj3D++TBiBCxbFnY1IiJSREGGfBpQN8t2HWBd9p3MrB3wIHCJc25nTgdyzg1zzjVzzjUrrdndwVi1ygf8xo0wdSqcdFLYFYmISBEFGfKfAw3MrJ6ZlQF6ABOy7mBmpwEv4AP+lwBrkbysWOEDfutWmDYNzjwz7IpERCQGAgt559we4HZgCrAEeMM5t8jMBpnZJZHdngYqAW+a2QIzm5DL4SRI6elQpgzMmAGnnx52NSIiEiOBLobjnJsETMr22IAs99sF+fqSj40boVYtOOMMWLrUr0svIiJJQ8vapqoFC+DEE+Ff//LbCngRkaSjkE9F8+dDmzZQvjxccEHY1YiISEAU8qnms8+gbVt/udjZs6F+/bArEhGRgCjkU8mGDf5qcoce6gP+mGPCrkhERAKkgdhUcsgh8O9/+1Xsjjwy7GpERCRgCvlUMHWqn1jXpg307Bl2NSIiUkwU8snu/fehWzdo1gxatwbLabVhERFJRhqTT2bjxsGll8Ipp8DEiQp4EZEUo5BPVm+8AVdc4Vew++gjqFEj7IpERKSYKeST1dSp0LKl/1q1atjViIhICDQmn2x+/90vcvPCC7BzJ1SoEHZFIiISErXkk8kLL8DJJ8P69VCypAJeRCTFKeSTxXPPwS23QMOGUL162NWIiEgcUMgng2eegX79oGtXeOcdKFcu7IpERCQOKOQT3YgR8Je/QPfufkZ9mTJhVyQiInFCE+8SXdeusHo1PPywLhcrIiIHUEs+ETkHI0f62fM1asDAgQp4ERE5iEI+0TgH99wDffrASy+FXY2IiMQxNf8SSWamn2A3dCjccQfcfHPYFYmISBxTSz5RZGb6UB86FP78Z3j2Wa1FLyIieVLIJ4rVq+Htt+HBB2HwYAW8iIjkS9318S4zE0qUgHr1YNEiOOKIsCsSEZEEoZZ8PNu9G668Eh57zG8r4EVEpAAU8vFq505/qdi33oJKlcKuRkREEpC66+NRRgZcdhlMmgTPPw+33RZ2RSIikoAU8vHGOb+K3dSpMGwY3HRT2BWJiEiCUsjHGzPo0QOuugquuy7sakREJIEp5OPFtm2wcCGcey707h12NSIikgQ08S4ebNkC7dtDp06waVPY1YiISJJQSz5smzbBBRfAN9/4mfQ1a4ZdkYiIJAmFfJg2bIB27WDZMhg/Hi68MOyKREQkiSjkwzRsGCxfDu+958NeREQkhjQmH6b774f58xXwIiISCIV8cVu9Glq3hh9+8GvSN2oUdkUiIpKkFPLF6fvv4fzz4auv/Hi8iIhIgDQmX1yWL4c2bWDHDpg+HZo2DbsiERFJcgr54rB8uW/B797tA/7UU8OuSEREUoC664vDoYdCs2Ywc6YCXkREio1a8kFasgSOPhqqVoUJE8KuRkREUoxa8kH5/HM46yxdJlZEREKjkA/Cp5/6c9+rV4dHHgm7GhERSVEK+VibPduvRX/YYTBrFhxzTNgViYhIitKYfCzt2uWvAV+3rp9Ff8QRYVckIiIpTCEfS2XK+HXoa9XyLXkREZEQqbs+FiZOhAEDwDlo3FgBLyIicUEhX1TvvAPdusHkyZCREXY1IiIi+ynki2LsWOjeHc44Az76CMqXD7siERGR/RTyhfXqq3D11XD22TBlil/wRkREJI4o5AurdGl/LvykSVC5ctjViIiIHEQhX1CrV/uvPXrABx9AxYrh1iMiIpILhXxBPPssHH88fPaZ3zYLtx4REZE8KOSjNXgw3HUXdO4Mp58edjUiIiL5UshH47HH4N57fRf92LF+0RsREZE4p5DPz+TJfqGba66BUaP8hDsREZEEoGVt89OxI4we7c+HL1ky7GpERESippZ8TpyDRx+F5cv95LqrrlLAi4hIwlHIZ5eZCbfdBgMHwpgxYVcjIiJSaOquzyozE26+GYYPh7/+FR5+OOyKRERECk0t+X327oU+fXzAP/QQ/O1vOg9eREQSmlry++zcCStXwqBBasGLSNLZvXs3aWlpZOhqmXGrXLly1KlTh9IxPItLIb9rl79VqgTTp+sceBFJSmlpaVSuXJljjjkGUy9l3HHOsWnTJtLS0qhXr17Mjpva3fU7d8Lll/tV7PbuVcCLSNLKyMigZs2aCvg4ZWbUrFkz5j0tqRvyv/8OXbvCxIlw5ZU6RU5Ekp4CPr4F8fNJzZDfsQMuvthfB374cLj11rArEhFJCePGjcPMWLp06f7HZs6cSefOnQ/Yr3fv3rz11luAn09w33330aBBA0466SSaN2/O5MmTi1zLU089Rf369TnhhBOYMmVKjvtMmzaNpk2b0qRJE8455xxWrFgBwOrVq2nbti2nnHIKrVq1Ii0tbf/3vPzyyzRo0IAGDRrw8ssvA7B9+3aaNGmy/1arVi3uuuuuIr+H/KRmyN9wA8yYAS+/7O+LiEixGDNmDOeccw5jx46N+nsefvhh1q9fz7fffsu3337LxIkT2b59e5HqWLx4MWPHjmXRokV88MEH/OlPf2Lv3r0H7Xfrrbfy2muvsWDBAnr27Mnjjz8OwJ///GeuvfZavv76awYMGMD9998PwObNmxk4cCBz585l3rx5DBw4kPT0dCpXrsyCBQv2344++mi6detWpPcQjdQM+Uce8ReaueaasCsREUkZv/76K5988gn//e9/ow75HTt28OKLL/Lcc89RtmxZAA477DC6d+9epFrGjx9Pjx49KFu2LPXq1aN+/frMmzfvoP3MjG3btgGwdetWjjzySMD/kdC2bVsAWrduzfjx4wGYMmUK7du3p0aNGlSvXp327dvzwQcfHHDM5cuX88svv3DuuecW6T1EI3Vm16en+5b7nXdCw4b+JiKSggZOXMTiddtiesxGR1bhkYsb57nPu+++S8eOHTn++OOpUaMGX375JU2bNs3ze1asWMFRRx1FlSpV8q3h7rvvZsaMGQc93qNHD+67774DHlu7di1nnnnm/u06deqwdu3ag753+PDhdOrUifLly1OlShXmzJkDwKmnnsrbb7/NnXfeybhx49i+fTubNm1i7dq11K1bN8/jjhkzhiuvvLJY5kikRshv3AgXXACLFkH79tA4719EERGJvTFjxuwfh+7RowdjxoyhadOmuYZdQUNwyJAhUe/rnIvq9YYMGcKkSZNo0aIFTz/9NP3792f48OE888wz3H777YwcOZLzzjuP2rVrU6pUqaiOO3bsWF599dWoay2K5A/5X36Bdu3gu+9g/HgFvIikvPxa3EHYtGkT06dP59tvv8XM2Lt3L2bG4MGDqVmzJunp6Qfsv3nzZmrVqkX9+vX58ccf2b59O5UrV87zNQrSkq9Tpw5r1qzZv52Wlra/K36fDRs2sHDhQlq0aAHAlVdeSceOHQE48sgjeeeddwA/DPH2229TtWpV6tSpw8yZMw84bqtWrfZvL1y4kD179nD66afn+V5iJbnH5Nevh1atYMUKeP99f9lYEREpdm+99RbXXnstq1ev5ocffmDNmjXUq1ePjz/+mAYNGrBu3TqWLFkC+JnrCxcupEmTJlSoUIEbbriBfv36sWvXLgDWr1/PqFGjDnqNIUOGHDC5bd8te8ADXHLJJYwdO5adO3eyatUqli9fTvPmzQ/Yp3r16mzdupXvvvsOgA8//JATTzwRgI0bN5KZmQn4WfrXX389AB06dGDq1Kmkp6eTnp7O1KlT6dChw/5jjhkzhquuuqqoH2fUkrslv2CBD/rJk+H888OuRkQkZY0ZM+agsL3ssssYPXo05557LqNGjaJPnz5kZGRQunRphg8fTtWqVQF4/PHHeeihh2jUqBHlypWjYsWKDBo0qEj1NG7cmO7du9OoUSNKlSrF0KFDKRlZL6VTp04MHz6cI488khdffJHLLruMEiVKUL16dUaMGAH40/7uv/9+zIzzzjuPoUOHAlCjRg0efvhhzjjjDAAGDBhAjRo19r/uG2+8waRJk4pUe0FYTuMH8azG0Se69g+M4PWbW+a+086dEJmFyZYtUK1a8RQnIhKnlixZsr8VKvErp5+TmX3hnGtWmOMlX3f9ypXQqBGMG+e3FfAiIpKikqu7ftkyaNvWL1l79NFhVyMiIhKq5An5xYuhTRvIzISZM+Hkk8OuSEREJFTJ0V2/bxa9mQJeRCQXiTYHK9UE8fNJjpA//HC46y6YNcuPx4uIyAHKlSvHpk2bFPRxat/15MuVKxfT4yZ2d/28eVCxol/g5oEHwq5GRCRu1alTh7S0NDZs2BB2KZKLcuXKUadOnZgeM9CQN7OOwLNASWC4c+5v2Z4vC7wCnA5sAq50zv0Q1cE/+QQuvNB3zX/8se+qFxGRHJUuXZp69eqFXYYUs8C6682sJDAUuBBoBFxlZtn70m8A0p1z9YEhwN+jOfaJ330FHTrAEUfA668r4EVERHIQ5Jh8c2CFc+5759wuYCzQJds+XYCXI/ffAtpaPlckKJ/xG/c/19+fIjdzJsS4a0NERCRZBBnytYE1WbbTIo/luI9zbg+wFaiZ10Grbt/MT4fWgRkzfEteREREchTkmHxOLfLs0zqj2Qcz6wv0jWz+eszalZs47LCNRaxP8lYL0GccLH3GwdNnHDx9xsE7obDfGGTIpwF1s2zXAdblsk+amZUCqgKbsx/IOTcMGLZv28zmF3YdX4mOPuPg6TMOnj7j4OkzDp6ZzS/s9wbZXf850MDM6plZGaAHMCHbPhOA6yL3LwemO53EKSIiEhOBteSdc3vM7HZgCv4UuhHOuUVmNgiY75ybAPwXeNXMVuBb8D2CqkdERCTVBHqevHNuEjAp22MDstzPAK4oxKGH5b+LFJE+4+DpMw6ePuPg6TMOXqE/44S7nryIiIhEJznWrhcREZGDxG3Im1lHM1tmZivM7L4cni9rZq9Hnp9rZscUf5WJLYrPuL+ZLTazr81smpkdHUadiS6/zznLfpebmTMzzVQuoGg+YzPrHvl9XmRmo4u7xkQXxf8XR/3/9u4/1uq6juP481VpQiplzJarSc4fhYRE1Ci3jDBntKicEx1oNK1Jv6ZFfzTcstUfTvOPSA3NOWQzRzgtZjlzDcUxrspCfsQsC5lzc8EasabQHL764/NhHW4XzvcA995zv7we23c753u+P973fc/O+34/53vfH0lrJG2snxlzRiPOsUzSfZJ2Stp6iNclaWn9HWyWNL3rQW333UK5Ue/vwFnAicAmYPKgbb4BLKuPrwRWjnbcY2lpmONZwPj6eFFyPDx5rtudAqwFBoAZox33WFoavpfPATYC76rPTx/tuMfS0jDH9wCL6uPJwI7RjnusLcCngOnA1kO8Pgd4jNJjZibwTLdja3dBaQAABdBJREFU9uuV/LC0xI2DdM2x7TW2X69PByi9DqI3Td7LAD8GbgX2jWRwLdEkx18D7rS9G8D2zhGOcaxrkmMDp9bHE/j/vijRhe21DNErpsMXgRUuBoB3Sjps69d+LfLD0hI3DtIkx52upfwFGb3pmmdJHwHeb/vRkQysRZq8l88FzpW0TtJAnSEzmmuS45uBBZJeofxX1bdHJrTjSq+f2307n/wxa4kbh9Q4f5IWADOAi4Y1onY6bJ4lvYUyA+PCkQqohZq8l99GGbL/NGVE6mlJU2z/a5hja4smOb4KWG77dkmfoPRAmWL7zeEP77jRc93r1yv5XlricriWuHFITXKMpIuBJcBc2/8ZodjapFueTwGmAE9K2kH5nm11br7rSdPPi9/afsP2S8BfKEU/mmmS42uBXwPYXg+cROlrH8dOo8/tTv1a5NMSd/h1zXEdRr6bUuDzHeaROWyebe+xPdH2JNuTKPc+zLV9xL2qj0NNPi9+Q7mRFEkTKcP320c0yrGtSY5fBmYDSPoQpcjvGtEo2281cE29y34msMf2q4fboS+H652WuMOuYY5vA04GVtV7Gl+2PXfUgh6DGuY5jkLDHD8OXCJpG7Af+L7tf45e1GNLwxx/D/ilpBspQ8gLc+HVG0kPUr5SmljvbfghcAKA7WWUex3mAH8DXge+2vWY+R1ERES0U78O10dERMRRSpGPiIhoqRT5iIiIlkqRj4iIaKkU+YiIiJZKkY8YYZL2S3q+Y5l0mG0nHWpGqh7P+WSdQWxTbe163hEc43pJ19THCyWd0fHavZImH+M4n5M0rcE+N0gaf7TnjmijFPmIkbfX9rSOZccInXe+7QsoEzvd1uvOtpfZXlGfLgTO6HjtOtvbjkmU/4vzLprFeQOQIh8xhBT5iD5Qr9iflvSnunxyiG3Ol/RsvfrfLOmcun5Bx/q7Jb21y+nWAmfXfWfX+b+31Lms317X31LnXt8s6ad13c2SFku6nDKXwQP1nOPqFfgMSYsk3doR80JJPz/CONfTMfmGpF9I2qAyH/yP6rrvUP7YWCNpTV13iaT1NY+rJJ3c5TwRrZUiHzHyxnUM1T9S1+0EPmt7OjAPWDrEftcDP7M9jVJkX6ntQ+cBF9b1+4H5Xc7/BWCLpJOA5cA82x+mdMBcJOk04MvA+banAj/p3Nn2Q8AGyhX3NNt7O15+CLis4/k8YOURxnkppR3tAUtszwCmAhdJmmp7KaV39yzbs2rL2puAi2suNwDf7XKeiNbqy7a2ES23txa6TicAd9TvoPdTeqsPth5YIul9wMO2X5Q0G/go8FxtPTyO8gfDUB6QtBfYQZkG9DzgJdt/ra/fD3wTuIMyr/29kn4HNJ4C1/YuSdtrX+0X6znW1eP2Euc7KO1Tp3esv0LS1ymfW+8FJgObB+07s65fV89zIiVvEcelFPmI/nAj8A/gAsoI277BG9j+laRngM8Dj0u6jjL15P22f9DgHPM7J76R9O6hNqp9yj9OmWzkSuBbwGd6+FlWAlcALwCP2LZKxW0cJ7AJuAW4E7hM0geAxcDHbO+WtJwyAcpgAp6wfVUP8Ua0VobrI/rDBODVOvf21ZSr2INIOgvYXoeoV1OGrf8IXC7p9LrNaZLObHjOF4BJks6uz68GnqrfYU+w/XvKTW1D3eH+b8o0uUN5GPgSZX7xlXVdT3HafoMy7D6zDvWfCrwG7JH0HuBzh4hlALjwwM8kabykoUZFIo4LKfIR/eEu4CuSBihD9a8Nsc08YKuk54EPAivqHe03AX+QtBl4gjKU3ZXtfZRZrFZJ2gK8CSyjFMxH6/GeoowyDLYcWHbgxrtBx90NbAPOtP1sXddznPW7/tuBxbY3ARuBPwP3Ub4COOAe4DFJa2zvotz5/2A9zwAlVxHHpcxCFxER0VK5ko+IiGipFPmIiIiWSpGPiIhoqRT5iIiIlkqRj4iIaKkU+YiIiJZKkY+IiGipFPmIiIiW+i+n5AQ447hB+gAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 576x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x576 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "         0.0       1.00      1.00      1.00     56864\n",
      "         1.0       0.92      0.82      0.86        98\n",
      "\n",
      "    accuracy                           1.00     56962\n",
      "   macro avg       0.96      0.91      0.93     56962\n",
      "weighted avg       1.00      1.00      1.00     56962\n",
      "\n"
     ]
    }
   ],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import pydot\n",
    "from pandas.io.parsers import read_csv\n",
    "from collections import Counter\n",
    "\n",
    "from p2 import regresionLogistica, prob\n",
    "\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.metrics import (confusion_matrix, precision_recall_curve, auc, roc_curve, recall_score, classification_report, f1_score,\n",
    "                             precision_recall_fscore_support)\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "from sklearn.utils.class_weight import compute_class_weight\n",
    "from scipy import linalg\n",
    "\n",
    "import seaborn as sns \n",
    "from matplotlib.colors import ListedColormap\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "from keras.models import Sequential\n",
    "from keras.layers import Activation, Dense, Dropout, BatchNormalization\n",
    "from keras.utils import plot_model\n",
    "from keras.optimizers import Adam\n",
    "\n",
    "from imblearn.under_sampling import NearMiss\n",
    "from imblearn.over_sampling import RandomOverSampler, SMOTE\n",
    "from imblearn.combine import SMOTETomek\n",
    "from imblearn.keras import BalancedBatchGenerator\n",
    "\n",
    "\n",
    "def get_csv(file_name):\n",
    "    datos = read_csv(file_name).values\n",
    "    datos = datos.astype(float)\n",
    "    return datos\n",
    "\n",
    "def get_X_dataset(datos):\n",
    "    X = datos[:,1:-2]\n",
    "    y = datos[:,-1]\n",
    "    return X, y\n",
    "\n",
    "def paint_classes(data): \n",
    "    n_class = pd.value_counts(data[:,-1], sort=True) \n",
    "    n_c = np.c_[np.unique(data[:,-1], return_counts=1)]\n",
    "    \n",
    "    frequencies = np.array(n_class)\n",
    "    freq_series = pd.Series.from_array(frequencies)\n",
    "\n",
    "    x_labels = ['Legímita (0.0)', 'Fraudulenta (1.0)']\n",
    "\n",
    "    plt.figure(figsize=(12, 8))\n",
    "    ax = freq_series.plot(kind='bar')\n",
    "    ax.set_title('Transacciones')\n",
    "    ax.set_xlabel('Clases')\n",
    "    ax.set_ylabel('Frequencia')\n",
    "    ax.set_xticklabels(x_labels, rotation=0)\n",
    "\n",
    "\n",
    "    def add_value_labels(ax, spacing=5):\n",
    "        for rect in ax.patches:\n",
    "            y_value = rect.get_height()\n",
    "            x_value = rect.get_x() + rect.get_width() / 2\n",
    "            space = spacing\n",
    "            va = 'bottom'\n",
    "            label = \"{:.1f}\".format(y_value)\n",
    "\n",
    "            ax.annotate(\n",
    "                label,                      \n",
    "                (x_value, y_value),        \n",
    "                xytext=(0, space),         \n",
    "                textcoords=\"offset points\", \n",
    "                ha='center',                \n",
    "                va=va)                      \n",
    "    add_value_labels(ax)\n",
    "\n",
    "    plt.savefig(\"det_class.png\")\n",
    "    \n",
    "def mostrarResultados(y_test,y_pred, epochs, nn_layers): \n",
    "    LABELS = ['Normal = 0.0', 'Fraud = 1.0']\n",
    "    confusion_matri = confusion_matrix(y_test, y_pred)\n",
    "    plt.figure(figsize=(8, 8))\n",
    "    sns.heatmap(confusion_matri, xticklabels=LABELS, yticklabels=LABELS, annot=True, fmt='d', cmap='YlGnBu', alpha=0.5, vmin=0)\n",
    "    plt.title(\"Matrix de Confusion\")\n",
    "    plt.ylabel('Verdaderos clase')\n",
    "    plt.xlabel('Clase Predicción')\n",
    "    plt.savefig('Figure_matrix_'+epochs+'_epochs_'+nn_layers+'.png')\n",
    "    plt.show()\n",
    "    print(classification_report(y_test,y_pred))\n",
    "\n",
    "def mostrarResultados_imblearn(y_test,y_pred): \n",
    "    LABELS = ['Normal = 0.0', 'Fraud = 1.0']\n",
    "    confusion_matri = confusion_matrix(y_test, y_pred)\n",
    "    plt.figure(figsize=(8, 8))\n",
    "    sns.heatmap(confusion_matri, xticklabels=LABELS, yticklabels=LABELS, annot=True, fmt='d', cmap='YlGnBu', alpha=0.5, vmin=0)\n",
    "    plt.title(\"Matrix de Confusion\")\n",
    "    plt.ylabel('Verdaderos clase')\n",
    "    plt.xlabel('Clase Predicción')\n",
    "    plt.show()\n",
    "    print(classification_report(y_test,y_pred))\n",
    "\n",
    "def model_reg_log(X_train,y_train,X_test, y_test):\n",
    "    theta = regresionLogistica(X_train,y_train)\n",
    "    result = prob(theta,X_test)\n",
    "    y_pred = np.where(result >= 0.5,1.,0.)\n",
    "    mostrarResultados_imblearn(y_test,y_pred)\n",
    "    return None\n",
    "\n",
    "def model_subSampling(X_train,y_train, X_test, y_test):\n",
    "    nm = NearMiss(version=1)\n",
    "    print('NearMiss - clase minoritaria = 1 (antes) {}'.format(Counter(y_train)))\n",
    "    X_train_res, y_train_res = nm.fit_sample(X_train, y_train)\n",
    "    print('NearMiss - clase minoritaria = 1 {}'.format(Counter(y_train_res)))\n",
    "    model_reg_log(X_train_res, y_train_res, X_test, y_test)\n",
    "    return None\n",
    "\n",
    "def model_oveSampling(X_train,y_train, X_test, y_test):\n",
    "    ros = RandomOverSampler()\n",
    "    print('RandomOverSampler clase mayoritaria (antes) = 0 {}'.format(Counter(y_train)))\n",
    "    X_train_res, y_train_res = ros.fit_sample(X_train, y_train)\n",
    "    print('RandomOverSampler clase mayoritaria = 0 {}'.format(Counter(y_train_res)))\n",
    "    model_reg_log(X_train_res, y_train_res, X_test, y_test)\n",
    "    return None\n",
    "\n",
    "def model_smote(X_train,y_train, X_test, y_test):\n",
    "    sm = SMOTE(random_state=27,sampling_strategy=1.0)\n",
    "    print('SMOTE clase mayoritaria (antes) = 0{}'.format(Counter(y_train)))\n",
    "    X_train_res, y_train_res = sm.fit_sample(X_train, y_train)\n",
    "    print('SMOTE clase mayoritaria (despues) = 0 {}'.format(Counter(y_train_res)))\n",
    "    model_reg_log(X_train_res, y_train_res, X_test, y_test)\n",
    "\n",
    "def model_over_sub_Sampling(X_train,y_train,  X_test, y_test):\n",
    "    smt = SMOTETomek()\n",
    "    print('SMOTETomek {}'.format(Counter(y_train)))\n",
    "\n",
    "\n",
    "    X_train_res, y_train_res = smt.fit_sample(X_train, y_train)\n",
    "    print('SMOTETomek {}'.format(Counter(y_train_res)))\n",
    "    #model_reg_log(X_train_res, y_train_res, X_test, y_test)\n",
    "    nn_modelo(X_train_res,y_train_res, X_test, y_test)\n",
    "    return None\n",
    "\n",
    "def obtiene_class_weights(y_train, smooth_factor):\n",
    "    \"\"\"\n",
    "    Devuelve los pesos de cada clase en función de las frecuencias de las muestras.\n",
    "    : param smooth_factor: factor que suaviza los pesos extremadamente desiguales\n",
    "    : param y: lista de etiquetas verdaderas (las etiquetas deben ser hashable)\n",
    "    : return: diccionario con el peso de cada clase factor suave\n",
    "    \"\"\"\n",
    "    counter = Counter(y_train)\n",
    "\n",
    "    if smooth_factor > 0:\n",
    "        p = max(counter.values()) * smooth_factor\n",
    "        for k in counter.keys():\n",
    "            counter[k] += p\n",
    "\n",
    "    majority = max(counter.values())\n",
    "\n",
    "    return {cls: float(majority / count) for cls, count in counter.items()}\n",
    "\n",
    "def compute_class_weights(y_train):\n",
    "    #class_weight = {0.0: 1,1.0: 512.87887}\n",
    "    \n",
    "    class_weight_list = compute_class_weight('balanced', np.unique(y_train), y_train)\n",
    "    class_weight = dict(zip(np.unique(y_train), class_weight_list))\n",
    "\n",
    "    return class_weight\n",
    "\n",
    "\n",
    "\n",
    "def plot_loss(history, epochs, nn_layers):\n",
    "    plt.figure(figsize=(8, 8))\n",
    "    plt.plot(history.history['loss'])\n",
    "    plt.plot(history.history['val_loss'])\n",
    "    plt.title('model loss')\n",
    "    plt.ylabel('loss')\n",
    "    plt.xlabel('epoch')\n",
    "    plt.legend(['train', 'test'], loc='upper right')\n",
    "    plt.savefig('Figure_loss_'+epochs+'_epochs_'+nn_layers+'.png')\n",
    "    plt.show()\n",
    "\n",
    "def plot_accu(history, epochs, nn_layers):\n",
    "    plt.figure(figsize=(8, 8))\n",
    "    plt.plot(history.history['accuracy'])\n",
    "    plt.plot(history.history['val_accuracy'])\n",
    "    plt.title('model accuracy')\n",
    "    plt.ylabel('accuracy')\n",
    "    plt.xlabel('epoch')\n",
    "    plt.legend(['train', 'test'], loc='lower right')\n",
    "    plt.savefig('Figure_acc_'+epochs+'_epochs_'+nn_layers+'.png')\n",
    "    plt.show()\n",
    "\n",
    "def plot_accu_recall(error_df, epochs, nn_layers):\n",
    "    precision, recall, th = precision_recall_curve(error_df.true_class, error_df.reconstruction_error)\n",
    "    plt.figure(figsize=(8, 8))\n",
    "    plt.plot(recall, precision, 'b', label='Precision-Recall curve')\n",
    "    plt.title('Recall vs Precision')\n",
    "    plt.xlabel('Recall')\n",
    "    plt.ylabel('Precision')\n",
    "    plt.savefig('Figure_accu_recall_'+epochs+'_epochs_'+nn_layers+'.png')\n",
    "    plt.show()\n",
    "\n",
    "def plot_tp_fn(error_df,  epochs, nn_layers):\n",
    "    fpr, tpr, thresholds = roc_curve(error_df.true_class, error_df.reconstruction_error)\n",
    "    roc_auc = auc(fpr, tpr)\n",
    "    plt.figure(figsize=(8, 8))\n",
    "    plt.title('Receiver Operating Characteristic')\n",
    "    plt.plot(fpr, tpr, label='AUC = %0.4f'% roc_auc)\n",
    "    plt.legend(loc='lower right')\n",
    "    plt.plot([0,1],[0,1],'r--')\n",
    "    plt.xlim([-0.001, 1])\n",
    "    plt.ylim([0, 1.001])\n",
    "    plt.ylabel('True Positive Rate')\n",
    "    plt.xlabel('False Positive Rate')\n",
    "    plt.savefig('Figure_tp_fn_'+epochs+'_epochs_'+nn_layers+'.png')\n",
    "    plt.show()\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "def nn_model(X_train):    \n",
    "    model = Sequential()\n",
    "    model.add(Dense(units=14, input_shape=(X_train.shape[1],), activation='tanh'))\n",
    "    model.add(Dense(units=7, activation='tanh'))\n",
    "    model.add(Dense(units=3, activation='tanh'))\n",
    "    model.add(Dense(units=1, activation='sigmoid'))\n",
    "    model.compile(Adam(lr=0.01), loss='binary_crossentropy', metrics=['accuracy'])\n",
    "\n",
    "    return model\n",
    "\n",
    "\n",
    "def nn_model_train_pred(model, X_train, y_train, X_test, y_test, class_weight):\n",
    "    epochs = 60\n",
    "    nn_layers = \"14_7_3_1_nor_class_smooth_factor_1\"\n",
    "\n",
    "    #NÚMERO DE CAPAS | Precision | Recall | Epochs\n",
    "\n",
    "    #32-18-8-1 = 0.87      0.83 // \n",
    "    #32-18-8-1 = 0.72      0.85 // 73\n",
    "    #32-18-8-1 = 0.89      0.81 // 86 \n",
    "    #32_18_8_1 = 0.84      0.83 // 73 No normalizada\n",
    "    #32_18_8_1 = 0.75      0.84 // 73 Normalizada\n",
    "    \n",
    "    #32_16_1 = 0.84      0.83 // 100 No normalizada\n",
    "    #32_16_1 = 0.66      0.86 // Normalizada\n",
    "    \n",
    "    #30–7–7–1= 0.62      0.86 //\n",
    "    #30-7-7-1= 0.71      0.85 //\n",
    "    #30-7-7-1= 0.75      0.84 // 88 \n",
    "    #30_7_7_1 = 0.84      0.83 // 100 -------------- Sin normalizar\n",
    "    #30_7_7_1 = 0.76      0.84 // 100 -------------- Normalizado\n",
    "    \n",
    "    \n",
    "    #30–14-7–7–1= 0.84      0.83 // 50 epochs\n",
    "    #30–14-7–7–1= 0.91      0.68  // 100 epochs\n",
    "    #30-14-7-7-1= 0.72      0.85 // 75 epochs\n",
    "    #30-14-7-7-1= 0.88      0.83 // 70 epochs\n",
    "    #30-14-7-7-1= 0.75      0.84 // 63\n",
    "    #30-14-7-7-1= 0.77      0.84 // 56\n",
    "    #30-14-7-7-1= 0.67      0.85 // 43\n",
    "    #30-14-7-7-1= 0.89      0.83 // 40\n",
    "    \n",
    "\n",
    "    #------------------------------------------------\n",
    "\n",
    "    #14-7-3-1 =  0.65      0.83\n",
    "    #14-7-3-1 =  0.59      0.86 // 100\n",
    "    #14-7-3-1 =  0.24      0.88 // 96\n",
    "    #14-7-3-1 =  0.72      0.86 // 59 -----------\n",
    "    #14-7-3-1 =  0.84      0.83 // 60 ----------- Sin normalizar\n",
    "    #14-7-3-1 =  0.71      0.86 // 60 ----------- Normalizada\n",
    "\n",
    "    #14-7-7-1 =  0.74      0.82 // \n",
    "    \n",
    "    \n",
    "    #30-7-7-1 =  0.73      0.86 \n",
    "    #30-7-7-1 =  0.79      0.83 //40\n",
    "    #30-7-7-1 =  0.78      0.83 //46\n",
    "    #30-7-7-1 =  0.77      0.85 //55\n",
    "    #30-7-7-1 =  0.82      0.81 //50 ------------ Nor\n",
    "    #30-7-7-1 =  0.77      0.84 //100\n",
    "\n",
    "    \n",
    "    #32-16-1 = 0.55      0.86 // 55\n",
    "    #32-16-1 = 0.34      0.86 // 75\n",
    "    #32-16-1 = 0.75      0.85 // 78\n",
    "    #32-16-1 = 0.72      0.85 // 100 ------------\n",
    "\n",
    "    print(class_weight)\n",
    "\n",
    "    #55\n",
    "    #plot_model(model)          \n",
    "                                                \n",
    "    history = model.fit(X_train, y_train, batch_size=32, verbose=1,  epochs=epochs, validation_data=(X_test, y_test), class_weight=class_weight) #\n",
    "    \n",
    "    print(history.history.keys())\n",
    "    y_pred = model.predict_classes(X_test)\n",
    "    \n",
    "    epochs = str(epochs)\n",
    "\n",
    "    mse = np.mean(np.power(X_test - y_pred, 2), axis=1)\n",
    "    error_df = pd.DataFrame({'reconstruction_error': mse, 'true_class': y_test})\n",
    "    print(error_df.describe())\n",
    "    plot_accu_recall(error_df, epochs, nn_layers)\n",
    "    plot_tp_fn(error_df, epochs, nn_layers)\n",
    "    plot_loss(history, epochs, nn_layers)\n",
    "    plot_accu(history, epochs, nn_layers)\n",
    "    mostrarResultados(y_test,y_pred, epochs, nn_layers)\n",
    "\n",
    "    return None\n",
    "\n",
    "\n",
    "\n",
    "def main():\n",
    "    path = 'creditcard.csv'\n",
    "    datos = get_csv(path)\n",
    "        \n",
    "    #paint_classes(datos)\n",
    "\n",
    "    X,y = get_X_dataset(datos)\n",
    "\n",
    "    X_train , X_test , y_train , y_test = train_test_split(X, y, test_size=0.20, random_state=32, stratify=y)\n",
    "\n",
    "    ##Permite Normalizar los datos y llevar un preprocesamiento\n",
    "    \n",
    "    sc = MinMaxScaler(feature_range=(-1, 1))\n",
    "    X_train = sc.fit_transform(X_train)\n",
    "    X_test = sc.transform(X_test)\n",
    "\n",
    "    #entrena el modelo regresion logistica \n",
    "    #model_reg_log(X_train,y_train, X_test, y_test)\n",
    "\n",
    "    ##Usando los algoritmos de muestreo con imblearn\n",
    "\n",
    "    #model_subSampling(X_train,y_train, X_test, y_test)\n",
    "    #model_oveSampling(X_train,y_train, X_test, y_test)\n",
    "    #model_smote(X_train,y_train, X_test, y_test)\n",
    "    #model_over_sub_Sampling(X_train,y_train, X_test, y_test)\n",
    "\n",
    "\n",
    "    \n",
    "\n",
    "    #RED NEURONAL \n",
    "    #Se obtiene el Modelo\n",
    "    model = nn_model(X_train)\n",
    "    \n",
    "    #Se obtiene el class_weight (balanceo de clases) con la función \"obtiene_class_weights\" (manual) o con \"compute_class_weight\" usando Scikit-learn\n",
    "    \n",
    "    clas_weight_ = obtiene_class_weights(y_train, 1)\n",
    "    print(f'clas_weight manual :{clas_weight_}')\n",
    "    clas_weight = compute_class_weights(y_train)\n",
    "    print(f'clas_weight Scikit-learn :{clas_weight}')\n",
    "\n",
    "\n",
    "    #Entrena y predice el modelo\n",
    "    nn_model_train_pred(model, X_train,y_train, X_test, y_test, clas_weight_)\n",
    "\n",
    "\n",
    "    #print(f'X_train :{X_train.shape}')\n",
    "    #print(f'X_train :{y_train.shape}')\n",
    "    #print(f'X_train :{X_test.shape}')\n",
    "    #print(f'X_train :{y_test.shape}')\n",
    "\n",
    "\n",
    "main()\n"
   ]
  },
  {
   "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": "3.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}