H2O_AutoML_IPYNB/decision_tree_ensembles-checkpoint.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# imports\n",
    "import h2o \n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "from h2o.estimators.gbm import H2OGradientBoostingEstimator \n",
    "from h2o.estimators.random_forest import H2ORandomForestEstimator\n",
    "from h2o.grid.grid_search import H2OGridSearch"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# display matplotlib graphics in notebook\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Checking whether there is an H2O instance running at http://localhost:54321..... not found.\n",
      "Attempting to start a local H2O server...\n",
      "  Java Version: java version \"1.8.0_112\"; Java(TM) SE Runtime Environment (build 1.8.0_112-b16); Java HotSpot(TM) 64-Bit Server VM (build 25.112-b16, mixed mode)\n",
      "  Starting server from /Users/phall/anaconda/lib/python3.5/site-packages/h2o/backend/bin/h2o.jar\n",
      "  Ice root: /var/folders/tc/0ss1l73113j3wdyjsxmy1j2r0000gn/T/tmplpdyw_ys\n",
      "  JVM stdout: /var/folders/tc/0ss1l73113j3wdyjsxmy1j2r0000gn/T/tmplpdyw_ys/h2o_phall_started_from_python.out\n",
      "  JVM stderr: /var/folders/tc/0ss1l73113j3wdyjsxmy1j2r0000gn/T/tmplpdyw_ys/h2o_phall_started_from_python.err\n",
      "  Server is running at http://127.0.0.1:54321\n",
      "Connecting to H2O server at http://127.0.0.1:54321... successful.\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div style=\"overflow:auto\"><table style=\"width:50%\"><tr><td>H2O cluster uptime:</td>\n",
       "<td>02 secs</td></tr>\n",
       "<tr><td>H2O cluster version:</td>\n",
       "<td>3.12.0.1</td></tr>\n",
       "<tr><td>H2O cluster version age:</td>\n",
       "<td>29 days </td></tr>\n",
       "<tr><td>H2O cluster name:</td>\n",
       "<td>H2O_from_python_phall_r5f286</td></tr>\n",
       "<tr><td>H2O cluster total nodes:</td>\n",
       "<td>1</td></tr>\n",
       "<tr><td>H2O cluster free memory:</td>\n",
       "<td>3.556 Gb</td></tr>\n",
       "<tr><td>H2O cluster total cores:</td>\n",
       "<td>8</td></tr>\n",
       "<tr><td>H2O cluster allowed cores:</td>\n",
       "<td>8</td></tr>\n",
       "<tr><td>H2O cluster status:</td>\n",
       "<td>accepting new members, healthy</td></tr>\n",
       "<tr><td>H2O connection url:</td>\n",
       "<td>http://127.0.0.1:54321</td></tr>\n",
       "<tr><td>H2O connection proxy:</td>\n",
       "<td>None</td></tr>\n",
       "<tr><td>H2O internal security:</td>\n",
       "<td>False</td></tr>\n",
       "<tr><td>Python version:</td>\n",
       "<td>3.5.2 final</td></tr></table></div>"
      ],
      "text/plain": [
       "--------------------------  ------------------------------\n",
       "H2O cluster uptime:         02 secs\n",
       "H2O cluster version:        3.12.0.1\n",
       "H2O cluster version age:    29 days\n",
       "H2O cluster name:           H2O_from_python_phall_r5f286\n",
       "H2O cluster total nodes:    1\n",
       "H2O cluster free memory:    3.556 Gb\n",
       "H2O cluster total cores:    8\n",
       "H2O cluster allowed cores:  8\n",
       "H2O cluster status:         accepting new members, healthy\n",
       "H2O connection url:         http://127.0.0.1:54321\n",
       "H2O connection proxy:\n",
       "H2O internal security:      False\n",
       "Python version:             3.5.2 final\n",
       "--------------------------  ------------------------------"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# start and connect to h2o server\n",
    "h2o.init()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# location of clean data file\n",
    "path = 'https://raw.githubusercontent.com/h2oai/app-consumer-loan/master/data/loan.csv'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# define input variable measurement levels \n",
    "# strings automatically parsed as enums (nominal)\n",
    "# numbers automatically parsed as numeric\n",
    "col_types = {'bad_loan': 'enum'}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Parse progress: |█████████████████████████████████████████████████████████| 100%\n"
     ]
    }
   ],
   "source": [
    "frame = h2o.import_file(path=path, col_types=col_types) # import from url"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Rows:163987\n",
      "Cols:15\n",
      "\n",
      "\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<table>\n",
       "<thead>\n",
       "<tr><th>       </th><th>loan_amnt         </th><th>term     </th><th>int_rate          </th><th>emp_length       </th><th>home_ownership  </th><th>annual_inc       </th><th>purpose           </th><th>addr_state  </th><th>dti               </th><th>delinq_2yrs        </th><th>revol_util        </th><th>total_acc         </th><th>bad_loan  </th><th>longest_credit_length  </th><th>verification_status  </th></tr>\n",
       "</thead>\n",
       "<tbody>\n",
       "<tr><td>type   </td><td>int               </td><td>enum     </td><td>real              </td><td>int              </td><td>enum            </td><td>real             </td><td>enum              </td><td>enum        </td><td>real              </td><td>int                </td><td>real              </td><td>int               </td><td>enum      </td><td>int                    </td><td>enum                 </td></tr>\n",
       "<tr><td>mins   </td><td>500.0             </td><td>         </td><td>5.42              </td><td>0.0              </td><td>                </td><td>1896.0           </td><td>                  </td><td>            </td><td>0.0               </td><td>0.0                </td><td>0.0               </td><td>1.0               </td><td>          </td><td>0.0                    </td><td>                     </td></tr>\n",
       "<tr><td>mean   </td><td>13074.169141456332</td><td>         </td><td>13.715904065566189</td><td>5.684352932995338</td><td>                </td><td>71915.67051974905</td><td>                  </td><td>            </td><td>15.881530121290167</td><td>0.22735700606252723</td><td>54.07917280242262 </td><td>24.579733834274574</td><td>          </td><td>14.854273655448333     </td><td>                     </td></tr>\n",
       "<tr><td>maxs   </td><td>35000.0           </td><td>         </td><td>26.060000000000002</td><td>10.0             </td><td>                </td><td>7141778.0        </td><td>                  </td><td>            </td><td>39.99             </td><td>29.0               </td><td>150.70000000000002</td><td>118.0             </td><td>          </td><td>65.0                   </td><td>                     </td></tr>\n",
       "<tr><td>sigma  </td><td>7993.556188734672 </td><td>         </td><td>4.391939870545809 </td><td>3.610663731100238</td><td>                </td><td>59070.91565491818</td><td>                  </td><td>            </td><td>7.5876682241925355</td><td>0.6941679229284191 </td><td>25.285366766770498</td><td>11.685190365910666</td><td>          </td><td>6.947732922546689      </td><td>                     </td></tr>\n",
       "<tr><td>zeros  </td><td>0                 </td><td>         </td><td>0                 </td><td>14248            </td><td>                </td><td>0                </td><td>                  </td><td>            </td><td>270               </td><td>139459             </td><td>1562              </td><td>0                 </td><td>          </td><td>11                     </td><td>                     </td></tr>\n",
       "<tr><td>missing</td><td>0                 </td><td>0        </td><td>0                 </td><td>5804             </td><td>0               </td><td>4                </td><td>0                 </td><td>0           </td><td>0                 </td><td>29                 </td><td>193               </td><td>29                </td><td>0         </td><td>29                     </td><td>0                    </td></tr>\n",
       "<tr><td>0      </td><td>5000.0            </td><td>36 months</td><td>10.65             </td><td>10.0             </td><td>RENT            </td><td>24000.0          </td><td>credit_card       </td><td>AZ          </td><td>27.650000000000002</td><td>0.0                </td><td>83.7              </td><td>9.0               </td><td>0         </td><td>26.0                   </td><td>verified             </td></tr>\n",
       "<tr><td>1      </td><td>2500.0            </td><td>60 months</td><td>15.27             </td><td>0.0              </td><td>RENT            </td><td>30000.0          </td><td>car               </td><td>GA          </td><td>1.0               </td><td>0.0                </td><td>9.4               </td><td>4.0               </td><td>1         </td><td>12.0                   </td><td>verified             </td></tr>\n",
       "<tr><td>2      </td><td>2400.0            </td><td>36 months</td><td>15.96             </td><td>10.0             </td><td>RENT            </td><td>12252.0          </td><td>small_business    </td><td>IL          </td><td>8.72              </td><td>0.0                </td><td>98.5              </td><td>10.0              </td><td>0         </td><td>10.0                   </td><td>not verified         </td></tr>\n",
       "<tr><td>3      </td><td>10000.0           </td><td>36 months</td><td>13.49             </td><td>10.0             </td><td>RENT            </td><td>49200.0          </td><td>other             </td><td>CA          </td><td>20.0              </td><td>0.0                </td><td>21.0              </td><td>37.0              </td><td>0         </td><td>15.0                   </td><td>verified             </td></tr>\n",
       "<tr><td>4      </td><td>5000.0            </td><td>36 months</td><td>7.9               </td><td>3.0              </td><td>RENT            </td><td>36000.0          </td><td>wedding           </td><td>AZ          </td><td>11.200000000000001</td><td>0.0                </td><td>28.3              </td><td>12.0              </td><td>0         </td><td>7.0                    </td><td>verified             </td></tr>\n",
       "<tr><td>5      </td><td>3000.0            </td><td>36 months</td><td>18.64             </td><td>9.0              </td><td>RENT            </td><td>48000.0          </td><td>car               </td><td>CA          </td><td>5.3500000000000005</td><td>0.0                </td><td>87.5              </td><td>4.0               </td><td>0         </td><td>4.0                    </td><td>verified             </td></tr>\n",
       "<tr><td>6      </td><td>5600.0            </td><td>60 months</td><td>21.28             </td><td>4.0              </td><td>OWN             </td><td>40000.0          </td><td>small_business    </td><td>CA          </td><td>5.55              </td><td>0.0                </td><td>32.6              </td><td>13.0              </td><td>1         </td><td>7.0                    </td><td>verified             </td></tr>\n",
       "<tr><td>7      </td><td>5375.0            </td><td>60 months</td><td>12.69             </td><td>0.0              </td><td>RENT            </td><td>15000.0          </td><td>other             </td><td>TX          </td><td>18.080000000000002</td><td>0.0                </td><td>36.5              </td><td>3.0               </td><td>1         </td><td>7.0                    </td><td>verified             </td></tr>\n",
       "<tr><td>8      </td><td>6500.0            </td><td>60 months</td><td>14.65             </td><td>5.0              </td><td>OWN             </td><td>72000.0          </td><td>debt_consolidation</td><td>AZ          </td><td>16.12             </td><td>0.0                </td><td>20.6              </td><td>23.0              </td><td>0         </td><td>13.0                   </td><td>not verified         </td></tr>\n",
       "<tr><td>9      </td><td>12000.0           </td><td>36 months</td><td>12.69             </td><td>10.0             </td><td>OWN             </td><td>75000.0          </td><td>debt_consolidation</td><td>CA          </td><td>10.78             </td><td>0.0                </td><td>67.10000000000001 </td><td>34.0              </td><td>0         </td><td>22.0                   </td><td>verified             </td></tr>\n",
       "</tbody>\n",
       "</table>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "frame.describe() # summarize data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# split into 40% training, 30% validation, and 30% test\n",
    "train, valid, test = frame.split_frame([0.4, 0.3])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "bad_loan\n",
      "['loan_amnt', 'term', 'int_rate', 'emp_length', 'home_ownership', 'annual_inc', 'purpose', 'addr_state', 'dti', 'delinq_2yrs', 'revol_util', 'total_acc', 'longest_credit_length', 'verification_status']\n"
     ]
    }
   ],
   "source": [
    "# assign target and inputs\n",
    "y = 'bad_loan'\n",
    "X = [name for name in frame.columns if name != y]\n",
    "print(y)\n",
    "print(X)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# set target to factor - for binary classification\n",
    "# just to be safe ... \n",
    "train[y] = train[y].asfactor()\n",
    "valid[y] = valid[y].asfactor()\n",
    "test[y] = test[y].asfactor()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "drf Model Build progress: |███████████████████████████████████████████████| 100%\n",
      "Model Details\n",
      "=============\n",
      "H2ORandomForestEstimator :  Distributed Random Forest\n",
      "Model Key:  rf_model\n",
      "\n",
      "\n",
      "ModelMetricsBinomial: drf\n",
      "** Reported on train data. **\n",
      "\n",
      "MSE: 0.1411167067612382\n",
      "RMSE: 0.3756550369171671\n",
      "LogLoss: 0.509813988268957\n",
      "Mean Per-Class Error: 0.37291342997920474\n",
      "AUC: 0.6762231231938096\n",
      "Gini: 0.35244624638761923\n",
      "Confusion Matrix (Act/Pred) for max f1 @ threshold = 0.1978395756200896: \n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div style=\"overflow:auto\"><table style=\"width:50%\"><tr><td><b></b></td>\n",
       "<td><b>0</b></td>\n",
       "<td><b>1</b></td>\n",
       "<td><b>Error</b></td>\n",
       "<td><b>Rate</b></td></tr>\n",
       "<tr><td>0</td>\n",
       "<td>34143.0</td>\n",
       "<td>19597.0</td>\n",
       "<td>0.3647</td>\n",
       "<td> (19597.0/53740.0)</td></tr>\n",
       "<tr><td>1</td>\n",
       "<td>4553.0</td>\n",
       "<td>7392.0</td>\n",
       "<td>0.3812</td>\n",
       "<td> (4553.0/11945.0)</td></tr>\n",
       "<tr><td>Total</td>\n",
       "<td>38696.0</td>\n",
       "<td>26989.0</td>\n",
       "<td>0.3677</td>\n",
       "<td> (24150.0/65685.0)</td></tr></table></div>"
      ],
      "text/plain": [
       "       0      1      Error    Rate\n",
       "-----  -----  -----  -------  -----------------\n",
       "0      34143  19597  0.3647   (19597.0/53740.0)\n",
       "1      4553   7392   0.3812   (4553.0/11945.0)\n",
       "Total  38696  26989  0.3677   (24150.0/65685.0)"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Maximum Metrics: Maximum metrics at their respective thresholds\n",
      "\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div style=\"overflow:auto\"><table style=\"width:50%\"><tr><td><b>metric</b></td>\n",
       "<td><b>threshold</b></td>\n",
       "<td><b>value</b></td>\n",
       "<td><b>idx</b></td></tr>\n",
       "<tr><td>max f1</td>\n",
       "<td>0.1978396</td>\n",
       "<td>0.3797195</td>\n",
       "<td>251.0</td></tr>\n",
       "<tr><td>max f2</td>\n",
       "<td>0.0952323</td>\n",
       "<td>0.5495621</td>\n",
       "<td>332.0</td></tr>\n",
       "<tr><td>max f0point5</td>\n",
       "<td>0.3333341</td>\n",
       "<td>0.3428376</td>\n",
       "<td>154.0</td></tr>\n",
       "<tr><td>max accuracy</td>\n",
       "<td>0.7024235</td>\n",
       "<td>0.8182995</td>\n",
       "<td>15.0</td></tr>\n",
       "<tr><td>max precision</td>\n",
       "<td>0.7024235</td>\n",
       "<td>0.578125</td>\n",
       "<td>15.0</td></tr>\n",
       "<tr><td>max recall</td>\n",
       "<td>0.0000028</td>\n",
       "<td>1.0</td>\n",
       "<td>399.0</td></tr>\n",
       "<tr><td>max specificity</td>\n",
       "<td>0.8535714</td>\n",
       "<td>0.9999814</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max absolute_mcc</td>\n",
       "<td>0.2647239</td>\n",
       "<td>0.2027558</td>\n",
       "<td>200.0</td></tr>\n",
       "<tr><td>max min_per_class_accuracy</td>\n",
       "<td>0.1944701</td>\n",
       "<td>0.6254373</td>\n",
       "<td>254.0</td></tr>\n",
       "<tr><td>max mean_per_class_accuracy</td>\n",
       "<td>0.1978396</td>\n",
       "<td>0.6270866</td>\n",
       "<td>251.0</td></tr></table></div>"
      ],
      "text/plain": [
       "metric                       threshold    value     idx\n",
       "---------------------------  -----------  --------  -----\n",
       "max f1                       0.19784      0.37972   251\n",
       "max f2                       0.0952323    0.549562  332\n",
       "max f0point5                 0.333334     0.342838  154\n",
       "max accuracy                 0.702423     0.818299  15\n",
       "max precision                0.702423     0.578125  15\n",
       "max recall                   2.76404e-06  1         399\n",
       "max specificity              0.853571     0.999981  0\n",
       "max absolute_mcc             0.264724     0.202756  200\n",
       "max min_per_class_accuracy   0.19447      0.625437  254\n",
       "max mean_per_class_accuracy  0.19784      0.627087  251"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Gains/Lift Table: Avg response rate: 18.19 %\n",
      "\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div style=\"overflow:auto\"><table style=\"width:50%\"><tr><td><b></b></td>\n",
       "<td><b>group</b></td>\n",
       "<td><b>cumulative_data_fraction</b></td>\n",
       "<td><b>lower_threshold</b></td>\n",
       "<td><b>lift</b></td>\n",
       "<td><b>cumulative_lift</b></td>\n",
       "<td><b>response_rate</b></td>\n",
       "<td><b>cumulative_response_rate</b></td>\n",
       "<td><b>capture_rate</b></td>\n",
       "<td><b>cumulative_capture_rate</b></td>\n",
       "<td><b>gain</b></td>\n",
       "<td><b>cumulative_gain</b></td></tr>\n",
       "<tr><td></td>\n",
       "<td>1</td>\n",
       "<td>0.0103829</td>\n",
       "<td>0.83</td>\n",
       "<td>5.4989535</td>\n",
       "<td>5.4989535</td>\n",
       "<td>1.0</td>\n",
       "<td>1.0</td>\n",
       "<td>0.0570950</td>\n",
       "<td>0.0570950</td>\n",
       "<td>449.8953537</td>\n",
       "<td>449.8953537</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2</td>\n",
       "<td>0.0200046</td>\n",
       "<td>0.8011264</td>\n",
       "<td>5.4989535</td>\n",
       "<td>5.4989535</td>\n",
       "<td>1.0</td>\n",
       "<td>1.0</td>\n",
       "<td>0.0529092</td>\n",
       "<td>0.1100042</td>\n",
       "<td>449.8953537</td>\n",
       "<td>449.8953537</td></tr>\n",
       "<tr><td></td>\n",
       "<td>3</td>\n",
       "<td>0.0300069</td>\n",
       "<td>0.7861778</td>\n",
       "<td>5.4989535</td>\n",
       "<td>5.4989535</td>\n",
       "<td>1.0</td>\n",
       "<td>1.0</td>\n",
       "<td>0.0550021</td>\n",
       "<td>0.1650063</td>\n",
       "<td>449.8953537</td>\n",
       "<td>449.8953537</td></tr>\n",
       "<tr><td></td>\n",
       "<td>4</td>\n",
       "<td>0.0440283</td>\n",
       "<td>0.77</td>\n",
       "<td>5.4989535</td>\n",
       "<td>5.4989535</td>\n",
       "<td>1.0</td>\n",
       "<td>1.0</td>\n",
       "<td>0.0771034</td>\n",
       "<td>0.2421097</td>\n",
       "<td>449.8953537</td>\n",
       "<td>449.8953537</td></tr>\n",
       "<tr><td></td>\n",
       "<td>5</td>\n",
       "<td>0.0527213</td>\n",
       "<td>0.76</td>\n",
       "<td>5.4989535</td>\n",
       "<td>5.4989535</td>\n",
       "<td>1.0</td>\n",
       "<td>1.0</td>\n",
       "<td>0.0478024</td>\n",
       "<td>0.2899121</td>\n",
       "<td>449.8953537</td>\n",
       "<td>449.8953537</td></tr>\n",
       "<tr><td></td>\n",
       "<td>6</td>\n",
       "<td>0.1022912</td>\n",
       "<td>0.71</td>\n",
       "<td>5.4989535</td>\n",
       "<td>5.4989535</td>\n",
       "<td>1.0</td>\n",
       "<td>1.0</td>\n",
       "<td>0.2725827</td>\n",
       "<td>0.5624948</td>\n",
       "<td>449.8953537</td>\n",
       "<td>449.8953537</td></tr>\n",
       "<tr><td></td>\n",
       "<td>7</td>\n",
       "<td>0.1500038</td>\n",
       "<td>0.6543496</td>\n",
       "<td>5.4989535</td>\n",
       "<td>5.4989535</td>\n",
       "<td>1.0</td>\n",
       "<td>1.0</td>\n",
       "<td>0.2623692</td>\n",
       "<td>0.8248640</td>\n",
       "<td>449.8953537</td>\n",
       "<td>449.8953537</td></tr>\n",
       "<tr><td></td>\n",
       "<td>8</td>\n",
       "<td>0.2004567</td>\n",
       "<td>0.1850000</td>\n",
       "<td>3.4712766</td>\n",
       "<td>4.9886079</td>\n",
       "<td>0.6312613</td>\n",
       "<td>0.9071922</td>\n",
       "<td>0.1751360</td>\n",
       "<td>1.0</td>\n",
       "<td>247.1276644</td>\n",
       "<td>398.8607883</td></tr>\n",
       "<tr><td></td>\n",
       "<td>9</td>\n",
       "<td>0.3046510</td>\n",
       "<td>0.12</td>\n",
       "<td>0.0</td>\n",
       "<td>3.2824447</td>\n",
       "<td>0.0</td>\n",
       "<td>0.5969217</td>\n",
       "<td>0.0</td>\n",
       "<td>1.0</td>\n",
       "<td>-100.0</td>\n",
       "<td>228.2444655</td></tr>\n",
       "<tr><td></td>\n",
       "<td>10</td>\n",
       "<td>0.4164269</td>\n",
       "<td>0.0900000</td>\n",
       "<td>0.0</td>\n",
       "<td>2.4013819</td>\n",
       "<td>0.0</td>\n",
       "<td>0.4366980</td>\n",
       "<td>0.0</td>\n",
       "<td>1.0</td>\n",
       "<td>-100.0</td>\n",
       "<td>140.1381933</td></tr>\n",
       "<tr><td></td>\n",
       "<td>11</td>\n",
       "<td>0.5270001</td>\n",
       "<td>0.0700000</td>\n",
       "<td>0.0</td>\n",
       "<td>1.8975329</td>\n",
       "<td>0.0</td>\n",
       "<td>0.3450716</td>\n",
       "<td>0.0</td>\n",
       "<td>1.0</td>\n",
       "<td>-100.0</td>\n",
       "<td>89.7532933</td></tr>\n",
       "<tr><td></td>\n",
       "<td>12</td>\n",
       "<td>0.6</td>\n",
       "<td>0.0554778</td>\n",
       "<td>0.0</td>\n",
       "<td>1.6666667</td>\n",
       "<td>0.0</td>\n",
       "<td>0.3030880</td>\n",
       "<td>0.0</td>\n",
       "<td>1.0</td>\n",
       "<td>-100.0</td>\n",
       "<td>66.6666667</td></tr>\n",
       "<tr><td></td>\n",
       "<td>13</td>\n",
       "<td>0.7478724</td>\n",
       "<td>0.0400000</td>\n",
       "<td>0.0</td>\n",
       "<td>1.3371265</td>\n",
       "<td>0.0</td>\n",
       "<td>0.2431602</td>\n",
       "<td>0.0</td>\n",
       "<td>1.0</td>\n",
       "<td>-100.0</td>\n",
       "<td>33.7126456</td></tr>\n",
       "<tr><td></td>\n",
       "<td>14</td>\n",
       "<td>0.8299155</td>\n",
       "<td>0.0300000</td>\n",
       "<td>0.0</td>\n",
       "<td>1.2049419</td>\n",
       "<td>0.0</td>\n",
       "<td>0.2191220</td>\n",
       "<td>0.0</td>\n",
       "<td>1.0</td>\n",
       "<td>-100.0</td>\n",
       "<td>20.4941940</td></tr>\n",
       "<tr><td></td>\n",
       "<td>15</td>\n",
       "<td>0.9039659</td>\n",
       "<td>0.0200000</td>\n",
       "<td>0.0</td>\n",
       "<td>1.1062364</td>\n",
       "<td>0.0</td>\n",
       "<td>0.2011722</td>\n",
       "<td>0.0</td>\n",
       "<td>1.0</td>\n",
       "<td>-100.0</td>\n",
       "<td>10.6236422</td></tr>\n",
       "<tr><td></td>\n",
       "<td>16</td>\n",
       "<td>1.0</td>\n",
       "<td>0.0</td>\n",
       "<td>0.0</td>\n",
       "<td>1.0</td>\n",
       "<td>0.0</td>\n",
       "<td>0.1818528</td>\n",
       "<td>0.0</td>\n",
       "<td>1.0</td>\n",
       "<td>-100.0</td>\n",
       "<td>0.0</td></tr></table></div>"
      ],
      "text/plain": [
       "    group    cumulative_data_fraction    lower_threshold    lift     cumulative_lift    response_rate    cumulative_response_rate    capture_rate    cumulative_capture_rate    gain     cumulative_gain\n",
       "--  -------  --------------------------  -----------------  -------  -----------------  ---------------  --------------------------  --------------  -------------------------  -------  -----------------\n",
       "    1        0.0103829                   0.83               5.49895  5.49895            1                1                           0.057095        0.057095                   449.895  449.895\n",
       "    2        0.0200046                   0.801126           5.49895  5.49895            1                1                           0.0529092       0.110004                   449.895  449.895\n",
       "    3        0.0300069                   0.786178           5.49895  5.49895            1                1                           0.0550021       0.165006                   449.895  449.895\n",
       "    4        0.0440283                   0.77               5.49895  5.49895            1                1                           0.0771034       0.24211                    449.895  449.895\n",
       "    5        0.0527213                   0.76               5.49895  5.49895            1                1                           0.0478024       0.289912                   449.895  449.895\n",
       "    6        0.102291                    0.71               5.49895  5.49895            1                1                           0.272583        0.562495                   449.895  449.895\n",
       "    7        0.150004                    0.65435            5.49895  5.49895            1                1                           0.262369        0.824864                   449.895  449.895\n",
       "    8        0.200457                    0.185              3.47128  4.98861            0.631261         0.907192                    0.175136        1                          247.128  398.861\n",
       "    9        0.304651                    0.12               0        3.28244            0                0.596922                    0               1                          -100     228.244\n",
       "    10       0.416427                    0.09               0        2.40138            0                0.436698                    0               1                          -100     140.138\n",
       "    11       0.527                       0.07               0        1.89753            0                0.345072                    0               1                          -100     89.7533\n",
       "    12       0.6                         0.0554778          0        1.66667            0                0.303088                    0               1                          -100     66.6667\n",
       "    13       0.747872                    0.04               0        1.33713            0                0.24316                     0               1                          -100     33.7126\n",
       "    14       0.829916                    0.03               0        1.20494            0                0.219122                    0               1                          -100     20.4942\n",
       "    15       0.903966                    0.02               0        1.10624            0                0.201172                    0               1                          -100     10.6236\n",
       "    16       1                           0                  0        1                  0                0.181853                    0               1                          -100     0"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "ModelMetricsBinomial: drf\n",
      "** Reported on validation data. **\n",
      "\n",
      "MSE: 0.1396948475554251\n",
      "RMSE: 0.37375773912445626\n",
      "LogLoss: 0.4467509809815742\n",
      "Mean Per-Class Error: 0.363237730268537\n",
      "AUC: 0.6854771212300304\n",
      "Gini: 0.3709542424600607\n",
      "Confusion Matrix (Act/Pred) for max f1 @ threshold = 0.20845103040337562: \n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div style=\"overflow:auto\"><table style=\"width:50%\"><tr><td><b></b></td>\n",
       "<td><b>0</b></td>\n",
       "<td><b>1</b></td>\n",
       "<td><b>Error</b></td>\n",
       "<td><b>Rate</b></td></tr>\n",
       "<tr><td>0</td>\n",
       "<td>26838.0</td>\n",
       "<td>13348.0</td>\n",
       "<td>0.3322</td>\n",
       "<td> (13348.0/40186.0)</td></tr>\n",
       "<tr><td>1</td>\n",
       "<td>3557.0</td>\n",
       "<td>5403.0</td>\n",
       "<td>0.397</td>\n",
       "<td> (3557.0/8960.0)</td></tr>\n",
       "<tr><td>Total</td>\n",
       "<td>30395.0</td>\n",
       "<td>18751.0</td>\n",
       "<td>0.344</td>\n",
       "<td> (16905.0/49146.0)</td></tr></table></div>"
      ],
      "text/plain": [
       "       0      1      Error    Rate\n",
       "-----  -----  -----  -------  -----------------\n",
       "0      26838  13348  0.3322   (13348.0/40186.0)\n",
       "1      3557   5403   0.397    (3557.0/8960.0)\n",
       "Total  30395  18751  0.344    (16905.0/49146.0)"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Maximum Metrics: Maximum metrics at their respective thresholds\n",
      "\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div style=\"overflow:auto\"><table style=\"width:50%\"><tr><td><b>metric</b></td>\n",
       "<td><b>threshold</b></td>\n",
       "<td><b>value</b></td>\n",
       "<td><b>idx</b></td></tr>\n",
       "<tr><td>max f1</td>\n",
       "<td>0.2084510</td>\n",
       "<td>0.3899534</td>\n",
       "<td>238.0</td></tr>\n",
       "<tr><td>max f2</td>\n",
       "<td>0.1050052</td>\n",
       "<td>0.5553456</td>\n",
       "<td>329.0</td></tr>\n",
       "<tr><td>max f0point5</td>\n",
       "<td>0.3080457</td>\n",
       "<td>0.3457447</td>\n",
       "<td>165.0</td></tr>\n",
       "<tr><td>max accuracy</td>\n",
       "<td>0.6340666</td>\n",
       "<td>0.8179302</td>\n",
       "<td>19.0</td></tr>\n",
       "<tr><td>max precision</td>\n",
       "<td>0.79</td>\n",
       "<td>1.0</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max recall</td>\n",
       "<td>0.0000089</td>\n",
       "<td>1.0</td>\n",
       "<td>399.0</td></tr>\n",
       "<tr><td>max specificity</td>\n",
       "<td>0.79</td>\n",
       "<td>1.0</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max absolute_mcc</td>\n",
       "<td>0.2131259</td>\n",
       "<td>0.2158844</td>\n",
       "<td>234.0</td></tr>\n",
       "<tr><td>max min_per_class_accuracy</td>\n",
       "<td>0.1959511</td>\n",
       "<td>0.6348214</td>\n",
       "<td>249.0</td></tr>\n",
       "<tr><td>max mean_per_class_accuracy</td>\n",
       "<td>0.1924923</td>\n",
       "<td>0.6367623</td>\n",
       "<td>252.0</td></tr></table></div>"
      ],
      "text/plain": [
       "metric                       threshold    value     idx\n",
       "---------------------------  -----------  --------  -----\n",
       "max f1                       0.208451     0.389953  238\n",
       "max f2                       0.105005     0.555346  329\n",
       "max f0point5                 0.308046     0.345745  165\n",
       "max accuracy                 0.634067     0.81793   19\n",
       "max precision                0.79         1         0\n",
       "max recall                   8.85873e-06  1         399\n",
       "max specificity              0.79         1         0\n",
       "max absolute_mcc             0.213126     0.215884  234\n",
       "max min_per_class_accuracy   0.195951     0.634821  249\n",
       "max mean_per_class_accuracy  0.192492     0.636762  252"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Gains/Lift Table: Avg response rate: 18.23 %\n",
      "\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div style=\"overflow:auto\"><table style=\"width:50%\"><tr><td><b></b></td>\n",
       "<td><b>group</b></td>\n",
       "<td><b>cumulative_data_fraction</b></td>\n",
       "<td><b>lower_threshold</b></td>\n",
       "<td><b>lift</b></td>\n",
       "<td><b>cumulative_lift</b></td>\n",
       "<td><b>response_rate</b></td>\n",
       "<td><b>cumulative_response_rate</b></td>\n",
       "<td><b>capture_rate</b></td>\n",
       "<td><b>cumulative_capture_rate</b></td>\n",
       "<td><b>gain</b></td>\n",
       "<td><b>cumulative_gain</b></td></tr>\n",
       "<tr><td></td>\n",
       "<td>1</td>\n",
       "<td>0.0100110</td>\n",
       "<td>0.5345077</td>\n",
       "<td>2.6310377</td>\n",
       "<td>2.6310377</td>\n",
       "<td>0.4796748</td>\n",
       "<td>0.4796748</td>\n",
       "<td>0.0263393</td>\n",
       "<td>0.0263393</td>\n",
       "<td>163.1037674</td>\n",
       "<td>163.1037674</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2</td>\n",
       "<td>0.0204493</td>\n",
       "<td>0.49</td>\n",
       "<td>2.3843371</td>\n",
       "<td>2.5051099</td>\n",
       "<td>0.4346979</td>\n",
       "<td>0.4567164</td>\n",
       "<td>0.0248884</td>\n",
       "<td>0.0512277</td>\n",
       "<td>138.4337145</td>\n",
       "<td>150.5109941</td></tr>\n",
       "<tr><td></td>\n",
       "<td>3</td>\n",
       "<td>0.0311724</td>\n",
       "<td>0.4600000</td>\n",
       "<td>2.4250767</td>\n",
       "<td>2.4775789</td>\n",
       "<td>0.4421252</td>\n",
       "<td>0.4516971</td>\n",
       "<td>0.0260045</td>\n",
       "<td>0.0772321</td>\n",
       "<td>142.5076664</td>\n",
       "<td>147.7578912</td></tr>\n",
       "<tr><td></td>\n",
       "<td>4</td>\n",
       "<td>0.0400033</td>\n",
       "<td>0.4385373</td>\n",
       "<td>2.2369883</td>\n",
       "<td>2.4244679</td>\n",
       "<td>0.4078341</td>\n",
       "<td>0.4420142</td>\n",
       "<td>0.0197545</td>\n",
       "<td>0.0969866</td>\n",
       "<td>123.6988253</td>\n",
       "<td>142.4467851</td></tr>\n",
       "<tr><td></td>\n",
       "<td>5</td>\n",
       "<td>0.0510316</td>\n",
       "<td>0.4200000</td>\n",
       "<td>2.0037617</td>\n",
       "<td>2.3335497</td>\n",
       "<td>0.3653137</td>\n",
       "<td>0.4254386</td>\n",
       "<td>0.0220982</td>\n",
       "<td>0.1190848</td>\n",
       "<td>100.3761696</td>\n",
       "<td>133.3549695</td></tr>\n",
       "<tr><td></td>\n",
       "<td>6</td>\n",
       "<td>0.1001302</td>\n",
       "<td>0.36</td>\n",
       "<td>1.8162249</td>\n",
       "<td>2.0798808</td>\n",
       "<td>0.3311231</td>\n",
       "<td>0.3791912</td>\n",
       "<td>0.0891741</td>\n",
       "<td>0.2082589</td>\n",
       "<td>81.6224894</td>\n",
       "<td>107.9880777</td></tr>\n",
       "<tr><td></td>\n",
       "<td>7</td>\n",
       "<td>0.1500834</td>\n",
       "<td>0.3183333</td>\n",
       "<td>1.7404683</td>\n",
       "<td>1.9669120</td>\n",
       "<td>0.3173116</td>\n",
       "<td>0.3585954</td>\n",
       "<td>0.0869420</td>\n",
       "<td>0.2952009</td>\n",
       "<td>74.0468341</td>\n",
       "<td>96.6912023</td></tr>\n",
       "<tr><td></td>\n",
       "<td>8</td>\n",
       "<td>0.2013185</td>\n",
       "<td>0.2850000</td>\n",
       "<td>1.4965154</td>\n",
       "<td>1.8471972</td>\n",
       "<td>0.2728356</td>\n",
       "<td>0.3367698</td>\n",
       "<td>0.0766741</td>\n",
       "<td>0.371875</td>\n",
       "<td>49.6515357</td>\n",
       "<td>84.7197165</td></tr>\n",
       "<tr><td></td>\n",
       "<td>9</td>\n",
       "<td>0.3000041</td>\n",
       "<td>0.2381701</td>\n",
       "<td>1.3537316</td>\n",
       "<td>1.6848730</td>\n",
       "<td>0.2468041</td>\n",
       "<td>0.3071758</td>\n",
       "<td>0.1335938</td>\n",
       "<td>0.5054688</td>\n",
       "<td>35.3731637</td>\n",
       "<td>68.4872978</td></tr>\n",
       "<tr><td></td>\n",
       "<td>10</td>\n",
       "<td>0.4091483</td>\n",
       "<td>0.2000000</td>\n",
       "<td>1.1483417</td>\n",
       "<td>1.5417482</td>\n",
       "<td>0.2093587</td>\n",
       "<td>0.2810822</td>\n",
       "<td>0.1253348</td>\n",
       "<td>0.6308036</td>\n",
       "<td>14.8341748</td>\n",
       "<td>54.1748176</td></tr>\n",
       "<tr><td></td>\n",
       "<td>11</td>\n",
       "<td>0.5054735</td>\n",
       "<td>0.1700000</td>\n",
       "<td>0.9524095</td>\n",
       "<td>1.4294412</td>\n",
       "<td>0.1736375</td>\n",
       "<td>0.2606070</td>\n",
       "<td>0.0917411</td>\n",
       "<td>0.7225446</td>\n",
       "<td>-4.7590474</td>\n",
       "<td>42.9441229</td></tr>\n",
       "<tr><td></td>\n",
       "<td>12</td>\n",
       "<td>0.6095715</td>\n",
       "<td>0.14</td>\n",
       "<td>0.8041015</td>\n",
       "<td>1.3226505</td>\n",
       "<td>0.1465989</td>\n",
       "<td>0.2411376</td>\n",
       "<td>0.0837054</td>\n",
       "<td>0.80625</td>\n",
       "<td>-19.5898459</td>\n",
       "<td>32.2650461</td></tr>\n",
       "<tr><td></td>\n",
       "<td>13</td>\n",
       "<td>0.7188988</td>\n",
       "<td>0.1100000</td>\n",
       "<td>0.6829508</td>\n",
       "<td>1.2253674</td>\n",
       "<td>0.1245114</td>\n",
       "<td>0.2234015</td>\n",
       "<td>0.0746652</td>\n",
       "<td>0.8809152</td>\n",
       "<td>-31.7049159</td>\n",
       "<td>22.5367450</td></tr>\n",
       "<tr><td></td>\n",
       "<td>14</td>\n",
       "<td>0.8000855</td>\n",
       "<td>0.0804651</td>\n",
       "<td>0.5746237</td>\n",
       "<td>1.1593349</td>\n",
       "<td>0.1047619</td>\n",
       "<td>0.2113629</td>\n",
       "<td>0.0466518</td>\n",
       "<td>0.9275670</td>\n",
       "<td>-42.5376276</td>\n",
       "<td>15.9334860</td></tr>\n",
       "<tr><td></td>\n",
       "<td>15</td>\n",
       "<td>0.9202784</td>\n",
       "<td>0.0500000</td>\n",
       "<td>0.4410693</td>\n",
       "<td>1.0655258</td>\n",
       "<td>0.0804131</td>\n",
       "<td>0.1942602</td>\n",
       "<td>0.0530134</td>\n",
       "<td>0.9805804</td>\n",
       "<td>-55.8930725</td>\n",
       "<td>6.5525830</td></tr>\n",
       "<tr><td></td>\n",
       "<td>16</td>\n",
       "<td>1.0</td>\n",
       "<td>0.0</td>\n",
       "<td>0.2435931</td>\n",
       "<td>1.0</td>\n",
       "<td>0.0444104</td>\n",
       "<td>0.1823139</td>\n",
       "<td>0.0194196</td>\n",
       "<td>1.0</td>\n",
       "<td>-75.6406899</td>\n",
       "<td>0.0</td></tr></table></div>"
      ],
      "text/plain": [
       "    group    cumulative_data_fraction    lower_threshold    lift      cumulative_lift    response_rate    cumulative_response_rate    capture_rate    cumulative_capture_rate    gain      cumulative_gain\n",
       "--  -------  --------------------------  -----------------  --------  -----------------  ---------------  --------------------------  --------------  -------------------------  --------  -----------------\n",
       "    1        0.010011                    0.534508           2.63104   2.63104            0.479675         0.479675                    0.0263393       0.0263393                  163.104   163.104\n",
       "    2        0.0204493                   0.49               2.38434   2.50511            0.434698         0.456716                    0.0248884       0.0512277                  138.434   150.511\n",
       "    3        0.0311724                   0.46               2.42508   2.47758            0.442125         0.451697                    0.0260045       0.0772321                  142.508   147.758\n",
       "    4        0.0400033                   0.438537           2.23699   2.42447            0.407834         0.442014                    0.0197545       0.0969866                  123.699   142.447\n",
       "    5        0.0510316                   0.42               2.00376   2.33355            0.365314         0.425439                    0.0220982       0.119085                   100.376   133.355\n",
       "    6        0.10013                     0.36               1.81622   2.07988            0.331123         0.379191                    0.0891741       0.208259                   81.6225   107.988\n",
       "    7        0.150083                    0.318333           1.74047   1.96691            0.317312         0.358595                    0.086942        0.295201                   74.0468   96.6912\n",
       "    8        0.201319                    0.285              1.49652   1.8472             0.272836         0.33677                     0.0766741       0.371875                   49.6515   84.7197\n",
       "    9        0.300004                    0.23817            1.35373   1.68487            0.246804         0.307176                    0.133594        0.505469                   35.3732   68.4873\n",
       "    10       0.409148                    0.2                1.14834   1.54175            0.209359         0.281082                    0.125335        0.630804                   14.8342   54.1748\n",
       "    11       0.505473                    0.17               0.95241   1.42944            0.173638         0.260607                    0.0917411       0.722545                   -4.75905  42.9441\n",
       "    12       0.609571                    0.14               0.804102  1.32265            0.146599         0.241138                    0.0837054       0.80625                    -19.5898  32.265\n",
       "    13       0.718899                    0.11               0.682951  1.22537            0.124511         0.223402                    0.0746652       0.880915                   -31.7049  22.5367\n",
       "    14       0.800085                    0.0804651          0.574624  1.15933            0.104762         0.211363                    0.0466518       0.927567                   -42.5376  15.9335\n",
       "    15       0.920278                    0.05               0.441069  1.06553            0.0804131        0.19426                     0.0530134       0.98058                    -55.8931  6.55258\n",
       "    16       1                           0                  0.243593  1                  0.0444104        0.182314                    0.0194196       1                          -75.6407  0"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Scoring History: \n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div style=\"overflow:auto\"><table style=\"width:50%\"><tr><td><b></b></td>\n",
       "<td><b>timestamp</b></td>\n",
       "<td><b>duration</b></td>\n",
       "<td><b>number_of_trees</b></td>\n",
       "<td><b>training_rmse</b></td>\n",
       "<td><b>training_logloss</b></td>\n",
       "<td><b>training_auc</b></td>\n",
       "<td><b>training_lift</b></td>\n",
       "<td><b>training_classification_error</b></td>\n",
       "<td><b>validation_rmse</b></td>\n",
       "<td><b>validation_logloss</b></td>\n",
       "<td><b>validation_auc</b></td>\n",
       "<td><b>validation_lift</b></td>\n",
       "<td><b>validation_classification_error</b></td></tr>\n",
       "<tr><td></td>\n",
       "<td>2017-07-06 16:55:47</td>\n",
       "<td> 0.060 sec</td>\n",
       "<td>0.0</td>\n",
       "<td>nan</td>\n",
       "<td>nan</td>\n",
       "<td>nan</td>\n",
       "<td>nan</td>\n",
       "<td>nan</td>\n",
       "<td>nan</td>\n",
       "<td>nan</td>\n",
       "<td>nan</td>\n",
       "<td>nan</td>\n",
       "<td>nan</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2017-07-06 16:55:48</td>\n",
       "<td> 0.838 sec</td>\n",
       "<td>1.0</td>\n",
       "<td>0.5272065</td>\n",
       "<td>9.5523061</td>\n",
       "<td>0.5385123</td>\n",
       "<td>3.9421275</td>\n",
       "<td>0.8149911</td>\n",
       "<td>0.5239222</td>\n",
       "<td>9.4408334</td>\n",
       "<td>0.5440907</td>\n",
       "<td>1.3810098</td>\n",
       "<td>0.8176861</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2017-07-06 16:55:48</td>\n",
       "<td> 1.309 sec</td>\n",
       "<td>2.0</td>\n",
       "<td>0.5112687</td>\n",
       "<td>8.3906374</td>\n",
       "<td>0.5460430</td>\n",
       "<td>5.2545151</td>\n",
       "<td>0.8156205</td>\n",
       "<td>0.4530274</td>\n",
       "<td>4.7091783</td>\n",
       "<td>0.5739908</td>\n",
       "<td>1.6421862</td>\n",
       "<td>0.3630204</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2017-07-06 16:55:48</td>\n",
       "<td> 1.742 sec</td>\n",
       "<td>3.0</td>\n",
       "<td>0.4952656</td>\n",
       "<td>7.4130004</td>\n",
       "<td>0.5529911</td>\n",
       "<td>5.4633573</td>\n",
       "<td>0.8163962</td>\n",
       "<td>0.4279959</td>\n",
       "<td>3.2555983</td>\n",
       "<td>0.5883271</td>\n",
       "<td>1.9509051</td>\n",
       "<td>0.4201359</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2017-07-06 16:55:49</td>\n",
       "<td> 2.070 sec</td>\n",
       "<td>4.0</td>\n",
       "<td>0.4819920</td>\n",
       "<td>6.5583097</td>\n",
       "<td>0.5580873</td>\n",
       "<td>5.4861610</td>\n",
       "<td>0.8173751</td>\n",
       "<td>0.4149903</td>\n",
       "<td>2.5442681</td>\n",
       "<td>0.5992393</td>\n",
       "<td>1.8283482</td>\n",
       "<td>0.4693159</td></tr>\n",
       "<tr><td>---</td>\n",
       "<td>---</td>\n",
       "<td>---</td>\n",
       "<td>---</td>\n",
       "<td>---</td>\n",
       "<td>---</td>\n",
       "<td>---</td>\n",
       "<td>---</td>\n",
       "<td>---</td>\n",
       "<td>---</td>\n",
       "<td>---</td>\n",
       "<td>---</td>\n",
       "<td>---</td>\n",
       "<td>---</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2017-07-06 16:59:04</td>\n",
       "<td> 3 min 17.576 sec</td>\n",
       "<td>96.0</td>\n",
       "<td>0.3758194</td>\n",
       "<td>0.5150384</td>\n",
       "<td>0.6753829</td>\n",
       "<td>5.4989535</td>\n",
       "<td>0.3479638</td>\n",
       "<td>0.3738096</td>\n",
       "<td>0.4469836</td>\n",
       "<td>0.6851647</td>\n",
       "<td>2.5835355</td>\n",
       "<td>0.3284499</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2017-07-06 16:59:08</td>\n",
       "<td> 3 min 21.343 sec</td>\n",
       "<td>97.0</td>\n",
       "<td>0.3757490</td>\n",
       "<td>0.5129747</td>\n",
       "<td>0.6758249</td>\n",
       "<td>5.4989535</td>\n",
       "<td>0.3582096</td>\n",
       "<td>0.3737971</td>\n",
       "<td>0.4469313</td>\n",
       "<td>0.6852647</td>\n",
       "<td>2.6190820</td>\n",
       "<td>0.3197208</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2017-07-06 16:59:12</td>\n",
       "<td> 3 min 25.231 sec</td>\n",
       "<td>98.0</td>\n",
       "<td>0.3756916</td>\n",
       "<td>0.5123152</td>\n",
       "<td>0.6760040</td>\n",
       "<td>5.4989535</td>\n",
       "<td>0.3514806</td>\n",
       "<td>0.3738106</td>\n",
       "<td>0.4469392</td>\n",
       "<td>0.6851663</td>\n",
       "<td>2.6421861</td>\n",
       "<td>0.3594189</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2017-07-06 16:59:16</td>\n",
       "<td> 3 min 29.163 sec</td>\n",
       "<td>99.0</td>\n",
       "<td>0.3756809</td>\n",
       "<td>0.5108562</td>\n",
       "<td>0.6760677</td>\n",
       "<td>5.4989535</td>\n",
       "<td>0.3515110</td>\n",
       "<td>0.3737720</td>\n",
       "<td>0.4468107</td>\n",
       "<td>0.6853822</td>\n",
       "<td>2.6345490</td>\n",
       "<td>0.3634477</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2017-07-06 16:59:20</td>\n",
       "<td> 3 min 33.252 sec</td>\n",
       "<td>100.0</td>\n",
       "<td>0.3756550</td>\n",
       "<td>0.5098140</td>\n",
       "<td>0.6762231</td>\n",
       "<td>5.4989535</td>\n",
       "<td>0.3676639</td>\n",
       "<td>0.3737577</td>\n",
       "<td>0.4467510</td>\n",
       "<td>0.6854771</td>\n",
       "<td>2.6310377</td>\n",
       "<td>0.3439751</td></tr></table></div>"
      ],
      "text/plain": [
       "     timestamp            duration          number_of_trees    training_rmse        training_logloss    training_auc        training_lift      training_classification_error    validation_rmse      validation_logloss    validation_auc      validation_lift     validation_classification_error\n",
       "---  -------------------  ----------------  -----------------  -------------------  ------------------  ------------------  -----------------  -------------------------------  -------------------  --------------------  ------------------  ------------------  ---------------------------------\n",
       "     2017-07-06 16:55:47  0.060 sec         0.0                nan                  nan                 nan                 nan                nan                              nan                  nan                   nan                 nan                 nan\n",
       "     2017-07-06 16:55:48  0.838 sec         1.0                0.5272064781671632   9.552306087970459   0.5385123082269838  3.942127514488832  0.8149911013616986               0.5239221638959581   9.440833365937703     0.5440906939539178  1.3810098182059656  0.8176860782159281\n",
       "     2017-07-06 16:55:48  1.309 sec         2.0                0.5112687190750738   8.390637366362958   0.5460430130291352  5.25451506500498   0.8156204787166536               0.45302742332373747  4.7091782611846735    0.5739908435262636  1.6421862427271605  0.3630203882309852\n",
       "     2017-07-06 16:55:48  1.742 sec         3.0                0.4952655648222379   7.413000399034582   0.5529910840467354  5.463357338171566  0.816396242527754                0.4279958734962778   3.255598277354007     0.5883271067993651  1.950905071330858   0.42013592153990154\n",
       "     2017-07-06 16:55:49  2.070 sec         4.0                0.48199202676217856  6.558309662136065   0.558087323687655   5.486161023000483  0.8173751174727102               0.4149902521945344   2.544268085341148     0.5992392628740641  1.8283482142857144  0.46931591584259147\n",
       "---  ---                  ---               ---                ---                  ---                 ---                 ---                ---                              ---                  ---                   ---                 ---                 ---\n",
       "     2017-07-06 16:59:04  3 min 17.576 sec  96.0               0.3758194382111227   0.515038400972803   0.6753829150882744  5.498953537044788  0.3479637664611403               0.3738096362901182   0.4469836465831731    0.6851646567790133  2.5835355201863357  0.3284499247141171\n",
       "     2017-07-06 16:59:08  3 min 21.343 sec  97.0               0.37574899329735434  0.5129747030250197  0.6758249306031879  5.498953537044788  0.35820963690340263              0.3737971160768941   0.44693129416994126   0.6852646993933565  2.619081982107912   0.3197208318072681\n",
       "     2017-07-06 16:59:12  3 min 25.231 sec  98.0               0.37569155170885016  0.5123151622977633  0.6760040358341318  5.498953537044788  0.3514805511151709               0.3738106364803329   0.4469391523650779    0.6851663092512673  2.6421861389372823  0.35941887437431325\n",
       "     2017-07-06 16:59:16  3 min 29.163 sec  99.0               0.37568089106242836  0.5108561907899344  0.67606772870259    5.498953537044788  0.35151099946715386              0.37377200095221874  0.4468107090854487    0.6853822012796745  2.6345490016872892  0.3634476864851666\n",
       "     2017-07-06 16:59:20  3 min 33.252 sec  100.0              0.3756550369171671   0.509813988268957   0.6762231231938096  5.498953537044788  0.36766385019410824              0.37375773912445626  0.4467509809815742    0.6854771212300304  2.631037674216028   0.343975094616042"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "See the whole table with table.as_data_frame()\n",
      "Variable Importances: \n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div style=\"overflow:auto\"><table style=\"width:50%\"><tr><td><b>variable</b></td>\n",
       "<td><b>relative_importance</b></td>\n",
       "<td><b>scaled_importance</b></td>\n",
       "<td><b>percentage</b></td></tr>\n",
       "<tr><td>addr_state</td>\n",
       "<td>107961.8281250</td>\n",
       "<td>1.0</td>\n",
       "<td>0.1764453</td></tr>\n",
       "<tr><td>int_rate</td>\n",
       "<td>74064.7578125</td>\n",
       "<td>0.6860273</td>\n",
       "<td>0.1210463</td></tr>\n",
       "<tr><td>dti</td>\n",
       "<td>58974.6875000</td>\n",
       "<td>0.5462550</td>\n",
       "<td>0.0963841</td></tr>\n",
       "<tr><td>revol_util</td>\n",
       "<td>57337.4843750</td>\n",
       "<td>0.5310903</td>\n",
       "<td>0.0937084</td></tr>\n",
       "<tr><td>loan_amnt</td>\n",
       "<td>51505.1757812</td>\n",
       "<td>0.4770684</td>\n",
       "<td>0.0841765</td></tr>\n",
       "<tr><td>total_acc</td>\n",
       "<td>49588.0156250</td>\n",
       "<td>0.4593106</td>\n",
       "<td>0.0810432</td></tr>\n",
       "<tr><td>longest_credit_length</td>\n",
       "<td>47351.0390625</td>\n",
       "<td>0.4385906</td>\n",
       "<td>0.0773873</td></tr>\n",
       "<tr><td>annual_inc</td>\n",
       "<td>45514.7617188</td>\n",
       "<td>0.4215820</td>\n",
       "<td>0.0743862</td></tr>\n",
       "<tr><td>emp_length</td>\n",
       "<td>39675.9843750</td>\n",
       "<td>0.3675001</td>\n",
       "<td>0.0648437</td></tr>\n",
       "<tr><td>purpose</td>\n",
       "<td>34946.6718750</td>\n",
       "<td>0.3236947</td>\n",
       "<td>0.0571144</td></tr>\n",
       "<tr><td>home_ownership</td>\n",
       "<td>13704.7705078</td>\n",
       "<td>0.1269409</td>\n",
       "<td>0.0223981</td></tr>\n",
       "<tr><td>term</td>\n",
       "<td>11918.3583984</td>\n",
       "<td>0.1103942</td>\n",
       "<td>0.0194785</td></tr>\n",
       "<tr><td>delinq_2yrs</td>\n",
       "<td>11131.6201172</td>\n",
       "<td>0.1031070</td>\n",
       "<td>0.0181927</td></tr>\n",
       "<tr><td>verification_status</td>\n",
       "<td>8196.1113281</td>\n",
       "<td>0.0759168</td>\n",
       "<td>0.0133952</td></tr></table></div>"
      ],
      "text/plain": [
       "variable               relative_importance    scaled_importance    percentage\n",
       "---------------------  ---------------------  -------------------  ------------\n",
       "addr_state             107962                 1                    0.176445\n",
       "int_rate               74064.8                0.686027             0.121046\n",
       "dti                    58974.7                0.546255             0.0963841\n",
       "revol_util             57337.5                0.53109              0.0937084\n",
       "loan_amnt              51505.2                0.477068             0.0841765\n",
       "total_acc              49588                  0.459311             0.0810432\n",
       "longest_credit_length  47351                  0.438591             0.0773873\n",
       "annual_inc             45514.8                0.421582             0.0743862\n",
       "emp_length             39676                  0.3675               0.0648437\n",
       "purpose                34946.7                0.323695             0.0571144\n",
       "home_ownership         13704.8                0.126941             0.0223981\n",
       "term                   11918.4                0.110394             0.0194785\n",
       "delinq_2yrs            11131.6                0.103107             0.0181927\n",
       "verification_status    8196.11                0.0759168            0.0133952"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": []
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# random forest\n",
    "# random forest is often best guess model with little tuning\n",
    "\n",
    "# initialize rf model\n",
    "rf_model = H2ORandomForestEstimator(\n",
    "    ntrees=500,                      # Up to 500 decision trees in the forest \n",
    "    max_depth=30,                    # trees can grow to depth of 30\n",
    "    stopping_rounds=5,               # stop after validation error does not decrease for 5 iterations/new trees\n",
    "    score_each_iteration=True,       # score validation error on every iteration/new tree\n",
    "    model_id='rf_model')             # for easy lookup in flow\n",
    "\n",
    "# train rf model\n",
    "rf_model.train(\n",
    "    x=X,\n",
    "    y=y,\n",
    "    training_frame=train,\n",
    "    validation_frame=valid)\n",
    "\n",
    "# print model information\n",
    "rf_model\n",
    "\n",
    "# view detailed results at http://ip:port/flow/index.html"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.6762231231938096\n",
      "0.6854771212300304\n",
      "0.6872709152702606\n"
     ]
    }
   ],
   "source": [
    "# measure rf AUC\n",
    "print(rf_model.auc(train=True))\n",
    "print(rf_model.auc(valid=True))\n",
    "print(rf_model.model_performance(test_data=test).auc())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "gbm Grid Build progress: |████████████████████████████████████████████████| 100%\n"
     ]
    }
   ],
   "source": [
    "# GBM is often more accurate the RF, but requires more tuning\n",
    "# GBM with random hyperparameter search\n",
    "# train many different GBM models with random hyperparameters\n",
    "# and select best model based on validation error\n",
    "\n",
    "# define random grid search parameters\n",
    "hyper_parameters = {'ntrees':list(range(50, 500, 50)),\n",
    "                    'max_depth':list(range(2, 20, 2)),\n",
    "                    'sample_rate':[s/float(10) for s in range(1, 11)],\n",
    "                    'col_sample_rate':[s/float(10) for s in range(1, 11)]}\n",
    "\n",
    "# define search strategy\n",
    "search_criteria = {'strategy':'RandomDiscrete',\n",
    "                   'max_models':20,\n",
    "                   'max_runtime_secs':600}\n",
    "\n",
    "# initialize grid search\n",
    "gsearch = H2OGridSearch(H2OGradientBoostingEstimator,\n",
    "                        hyper_params=hyper_parameters,\n",
    "                        search_criteria=search_criteria)\n",
    "\n",
    "# execute training w/ grid search\n",
    "gsearch.train(x=X,\n",
    "              y=y,\n",
    "              training_frame=train,\n",
    "              validation_frame=valid)\n",
    "\n",
    "# view detailed results at http://ip:port/flow/index.html"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "     col_sample_rate max_depth ntrees sample_rate  \\\n",
      "0                0.7         2    350         0.6   \n",
      "1                0.6         2    300         0.7   \n",
      "2                0.1         4    350         0.7   \n",
      "3                0.6         4    400         0.7   \n",
      "4                0.1         4    100         0.1   \n",
      "5                0.7         6    400         1.0   \n",
      "6                0.9         6    300         1.0   \n",
      "7                0.7         6    200         0.6   \n",
      "8                0.7        14     50         0.5   \n",
      "9                0.7        14     50         0.4   \n",
      "10               0.6        12    100         1.0   \n",
      "11               0.8        10    200         0.8   \n",
      "12               0.3        10    350         0.5   \n",
      "13               0.3        18    100         0.3   \n",
      "14               0.5        10    350         0.4   \n",
      "15               0.5        14    150         0.8   \n",
      "16               1.0        14    250         0.3   \n",
      "17               0.5        18    150         1.0   \n",
      "18               0.9        14    400         0.1   \n",
      "19               0.6        18    300         0.5   \n",
      "\n",
      "                                                         model_ids  \\\n",
      "0   Grid_GBM_py_7_sid_95b0_model_python_1499374541716_409_model_10   \n",
      "1   Grid_GBM_py_7_sid_95b0_model_python_1499374541716_409_model_19   \n",
      "2   Grid_GBM_py_7_sid_95b0_model_python_1499374541716_409_model_12   \n",
      "3   Grid_GBM_py_7_sid_95b0_model_python_1499374541716_409_model_18   \n",
      "4    Grid_GBM_py_7_sid_95b0_model_python_1499374541716_409_model_3   \n",
      "5    Grid_GBM_py_7_sid_95b0_model_python_1499374541716_409_model_6   \n",
      "6    Grid_GBM_py_7_sid_95b0_model_python_1499374541716_409_model_7   \n",
      "7   Grid_GBM_py_7_sid_95b0_model_python_1499374541716_409_model_11   \n",
      "8   Grid_GBM_py_7_sid_95b0_model_python_1499374541716_409_model_13   \n",
      "9   Grid_GBM_py_7_sid_95b0_model_python_1499374541716_409_model_15   \n",
      "10  Grid_GBM_py_7_sid_95b0_model_python_1499374541716_409_model_17   \n",
      "11   Grid_GBM_py_7_sid_95b0_model_python_1499374541716_409_model_4   \n",
      "12   Grid_GBM_py_7_sid_95b0_model_python_1499374541716_409_model_0   \n",
      "13  Grid_GBM_py_7_sid_95b0_model_python_1499374541716_409_model_14   \n",
      "14   Grid_GBM_py_7_sid_95b0_model_python_1499374541716_409_model_8   \n",
      "15   Grid_GBM_py_7_sid_95b0_model_python_1499374541716_409_model_1   \n",
      "16   Grid_GBM_py_7_sid_95b0_model_python_1499374541716_409_model_5   \n",
      "17   Grid_GBM_py_7_sid_95b0_model_python_1499374541716_409_model_9   \n",
      "18   Grid_GBM_py_7_sid_95b0_model_python_1499374541716_409_model_2   \n",
      "19  Grid_GBM_py_7_sid_95b0_model_python_1499374541716_409_model_16   \n",
      "\n",
      "                logloss  \n",
      "0   0.43535916538419495  \n",
      "1   0.43552628550049044  \n",
      "2   0.43591010316745027  \n",
      "3    0.4378764205320232  \n",
      "4    0.4389729823351246  \n",
      "5   0.43943263947774525  \n",
      "6   0.43967532839734746  \n",
      "7    0.4433546955422222  \n",
      "8    0.4636062753540612  \n",
      "9    0.4651195098968491  \n",
      "10   0.4715523579286705  \n",
      "11   0.4736613692521811  \n",
      "12   0.5009143944796578  \n",
      "13    0.505212908874478  \n",
      "14   0.5259157880555237  \n",
      "15   0.5416826357189778  \n",
      "16    0.599503416719909  \n",
      "17   0.6836337251545191  \n",
      "18   0.7147135255493922  \n",
      "19   0.8093384055472133  \n",
      "Model Details\n",
      "=============\n",
      "H2OGradientBoostingEstimator :  Gradient Boosting Machine\n",
      "Model Key:  Grid_GBM_py_7_sid_95b0_model_python_1499374541716_409_model_10\n",
      "\n",
      "\n",
      "ModelMetricsBinomial: gbm\n",
      "** Reported on train data. **\n",
      "\n",
      "MSE: 0.13309978441072112\n",
      "RMSE: 0.36482843147255\n",
      "LogLoss: 0.4234491377496304\n",
      "Mean Per-Class Error: 0.336786004206415\n",
      "AUC: 0.7267849815001551\n",
      "Gini: 0.4535699630003103\n",
      "Confusion Matrix (Act/Pred) for max f1 @ threshold = 0.2192987852671643: \n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div style=\"overflow:auto\"><table style=\"width:50%\"><tr><td><b></b></td>\n",
       "<td><b>0</b></td>\n",
       "<td><b>1</b></td>\n",
       "<td><b>Error</b></td>\n",
       "<td><b>Rate</b></td></tr>\n",
       "<tr><td>0</td>\n",
       "<td>40366.0</td>\n",
       "<td>13374.0</td>\n",
       "<td>0.2489</td>\n",
       "<td> (13374.0/53740.0)</td></tr>\n",
       "<tr><td>1</td>\n",
       "<td>5182.0</td>\n",
       "<td>6763.0</td>\n",
       "<td>0.4338</td>\n",
       "<td> (5182.0/11945.0)</td></tr>\n",
       "<tr><td>Total</td>\n",
       "<td>45548.0</td>\n",
       "<td>20137.0</td>\n",
       "<td>0.2825</td>\n",
       "<td> (18556.0/65685.0)</td></tr></table></div>"
      ],
      "text/plain": [
       "       0      1      Error    Rate\n",
       "-----  -----  -----  -------  -----------------\n",
       "0      40366  13374  0.2489   (13374.0/53740.0)\n",
       "1      5182   6763   0.4338   (5182.0/11945.0)\n",
       "Total  45548  20137  0.2825   (18556.0/65685.0)"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Maximum Metrics: Maximum metrics at their respective thresholds\n",
      "\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div style=\"overflow:auto\"><table style=\"width:50%\"><tr><td><b>metric</b></td>\n",
       "<td><b>threshold</b></td>\n",
       "<td><b>value</b></td>\n",
       "<td><b>idx</b></td></tr>\n",
       "<tr><td>max f1</td>\n",
       "<td>0.2192988</td>\n",
       "<td>0.4216071</td>\n",
       "<td>211.0</td></tr>\n",
       "<tr><td>max f2</td>\n",
       "<td>0.1197947</td>\n",
       "<td>0.5742194</td>\n",
       "<td>303.0</td></tr>\n",
       "<tr><td>max f0point5</td>\n",
       "<td>0.3148313</td>\n",
       "<td>0.4050938</td>\n",
       "<td>139.0</td></tr>\n",
       "<tr><td>max accuracy</td>\n",
       "<td>0.4947981</td>\n",
       "<td>0.8216488</td>\n",
       "<td>48.0</td></tr>\n",
       "<tr><td>max precision</td>\n",
       "<td>0.7515726</td>\n",
       "<td>1.0</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max recall</td>\n",
       "<td>0.0204400</td>\n",
       "<td>1.0</td>\n",
       "<td>396.0</td></tr>\n",
       "<tr><td>max specificity</td>\n",
       "<td>0.7515726</td>\n",
       "<td>1.0</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max absolute_mcc</td>\n",
       "<td>0.2587407</td>\n",
       "<td>0.2680419</td>\n",
       "<td>178.0</td></tr>\n",
       "<tr><td>max min_per_class_accuracy</td>\n",
       "<td>0.1862467</td>\n",
       "<td>0.6624488</td>\n",
       "<td>239.0</td></tr>\n",
       "<tr><td>max mean_per_class_accuracy</td>\n",
       "<td>0.1875129</td>\n",
       "<td>0.6632140</td>\n",
       "<td>238.0</td></tr></table></div>"
      ],
      "text/plain": [
       "metric                       threshold    value     idx\n",
       "---------------------------  -----------  --------  -----\n",
       "max f1                       0.219299     0.421607  211\n",
       "max f2                       0.119795     0.574219  303\n",
       "max f0point5                 0.314831     0.405094  139\n",
       "max accuracy                 0.494798     0.821649  48\n",
       "max precision                0.751573     1         0\n",
       "max recall                   0.02044      1         396\n",
       "max specificity              0.751573     1         0\n",
       "max absolute_mcc             0.258741     0.268042  178\n",
       "max min_per_class_accuracy   0.186247     0.662449  239\n",
       "max mean_per_class_accuracy  0.187513     0.663214  238"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Gains/Lift Table: Avg response rate: 18.19 %\n",
      "\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div style=\"overflow:auto\"><table style=\"width:50%\"><tr><td><b></b></td>\n",
       "<td><b>group</b></td>\n",
       "<td><b>cumulative_data_fraction</b></td>\n",
       "<td><b>lower_threshold</b></td>\n",
       "<td><b>lift</b></td>\n",
       "<td><b>cumulative_lift</b></td>\n",
       "<td><b>response_rate</b></td>\n",
       "<td><b>cumulative_response_rate</b></td>\n",
       "<td><b>capture_rate</b></td>\n",
       "<td><b>cumulative_capture_rate</b></td>\n",
       "<td><b>gain</b></td>\n",
       "<td><b>cumulative_gain</b></td></tr>\n",
       "<tr><td></td>\n",
       "<td>1</td>\n",
       "<td>0.0100023</td>\n",
       "<td>0.5275738</td>\n",
       "<td>3.4902034</td>\n",
       "<td>3.4902034</td>\n",
       "<td>0.6347032</td>\n",
       "<td>0.6347032</td>\n",
       "<td>0.0349100</td>\n",
       "<td>0.0349100</td>\n",
       "<td>249.0203387</td>\n",
       "<td>249.0203387</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2</td>\n",
       "<td>0.0200046</td>\n",
       "<td>0.4845127</td>\n",
       "<td>2.9210575</td>\n",
       "<td>3.2056304</td>\n",
       "<td>0.5312024</td>\n",
       "<td>0.5829528</td>\n",
       "<td>0.0292172</td>\n",
       "<td>0.0641272</td>\n",
       "<td>192.1057511</td>\n",
       "<td>220.5630449</td></tr>\n",
       "<tr><td></td>\n",
       "<td>3</td>\n",
       "<td>0.0300069</td>\n",
       "<td>0.4532832</td>\n",
       "<td>2.6364846</td>\n",
       "<td>3.0159152</td>\n",
       "<td>0.4794521</td>\n",
       "<td>0.5484526</td>\n",
       "<td>0.0263709</td>\n",
       "<td>0.0904981</td>\n",
       "<td>163.6484573</td>\n",
       "<td>201.5915157</td></tr>\n",
       "<tr><td></td>\n",
       "<td>4</td>\n",
       "<td>0.0400091</td>\n",
       "<td>0.4308361</td>\n",
       "<td>2.6699637</td>\n",
       "<td>2.9294273</td>\n",
       "<td>0.4855403</td>\n",
       "<td>0.5327245</td>\n",
       "<td>0.0267057</td>\n",
       "<td>0.1172039</td>\n",
       "<td>166.9963742</td>\n",
       "<td>192.9427303</td></tr>\n",
       "<tr><td></td>\n",
       "<td>5</td>\n",
       "<td>0.0500114</td>\n",
       "<td>0.4111860</td>\n",
       "<td>2.6532242</td>\n",
       "<td>2.8741867</td>\n",
       "<td>0.4824962</td>\n",
       "<td>0.5226788</td>\n",
       "<td>0.0265383</td>\n",
       "<td>0.1437422</td>\n",
       "<td>165.3224157</td>\n",
       "<td>187.4186674</td></tr>\n",
       "<tr><td></td>\n",
       "<td>6</td>\n",
       "<td>0.1000076</td>\n",
       "<td>0.3457907</td>\n",
       "<td>2.2119725</td>\n",
       "<td>2.5431300</td>\n",
       "<td>0.4022533</td>\n",
       "<td>0.4624753</td>\n",
       "<td>0.1105902</td>\n",
       "<td>0.2543324</td>\n",
       "<td>121.1972479</td>\n",
       "<td>154.3129981</td></tr>\n",
       "<tr><td></td>\n",
       "<td>7</td>\n",
       "<td>0.1500038</td>\n",
       "<td>0.3017843</td>\n",
       "<td>1.8368916</td>\n",
       "<td>2.3077411</td>\n",
       "<td>0.3340438</td>\n",
       "<td>0.4196691</td>\n",
       "<td>0.0918376</td>\n",
       "<td>0.3461699</td>\n",
       "<td>83.6891605</td>\n",
       "<td>130.7741081</td></tr>\n",
       "<tr><td></td>\n",
       "<td>8</td>\n",
       "<td>0.2</td>\n",
       "<td>0.2697795</td>\n",
       "<td>1.5589299</td>\n",
       "<td>2.1205525</td>\n",
       "<td>0.2834957</td>\n",
       "<td>0.3856284</td>\n",
       "<td>0.0779406</td>\n",
       "<td>0.4241105</td>\n",
       "<td>55.8929885</td>\n",
       "<td>112.0552532</td></tr>\n",
       "<tr><td></td>\n",
       "<td>9</td>\n",
       "<td>0.3000076</td>\n",
       "<td>0.2212764</td>\n",
       "<td>1.3402077</td>\n",
       "<td>1.8604244</td>\n",
       "<td>0.2437205</td>\n",
       "<td>0.3383234</td>\n",
       "<td>0.1340310</td>\n",
       "<td>0.5581415</td>\n",
       "<td>34.0207735</td>\n",
       "<td>86.0424400</td></tr>\n",
       "<tr><td></td>\n",
       "<td>10</td>\n",
       "<td>0.4</td>\n",
       "<td>0.1847910</td>\n",
       "<td>1.0850554</td>\n",
       "<td>1.6665969</td>\n",
       "<td>0.1973203</td>\n",
       "<td>0.3030753</td>\n",
       "<td>0.1084973</td>\n",
       "<td>0.6666388</td>\n",
       "<td>8.5055387</td>\n",
       "<td>66.6596902</td></tr>\n",
       "<tr><td></td>\n",
       "<td>11</td>\n",
       "<td>0.5000076</td>\n",
       "<td>0.1553728</td>\n",
       "<td>0.9291884</td>\n",
       "<td>1.5191062</td>\n",
       "<td>0.1689755</td>\n",
       "<td>0.2762537</td>\n",
       "<td>0.0929259</td>\n",
       "<td>0.7595647</td>\n",
       "<td>-7.0811626</td>\n",
       "<td>51.9106216</td></tr>\n",
       "<tr><td></td>\n",
       "<td>12</td>\n",
       "<td>0.6</td>\n",
       "<td>0.1301437</td>\n",
       "<td>0.7610458</td>\n",
       "<td>1.3927724</td>\n",
       "<td>0.1383983</td>\n",
       "<td>0.2532795</td>\n",
       "<td>0.0760988</td>\n",
       "<td>0.8356635</td>\n",
       "<td>-23.8954207</td>\n",
       "<td>39.2772429</td></tr>\n",
       "<tr><td></td>\n",
       "<td>13</td>\n",
       "<td>0.6999924</td>\n",
       "<td>0.1070972</td>\n",
       "<td>0.6329490</td>\n",
       "<td>1.2842333</td>\n",
       "<td>0.1151035</td>\n",
       "<td>0.2335414</td>\n",
       "<td>0.0632901</td>\n",
       "<td>0.8989535</td>\n",
       "<td>-36.7051024</td>\n",
       "<td>28.4233304</td></tr>\n",
       "<tr><td></td>\n",
       "<td>14</td>\n",
       "<td>0.8</td>\n",
       "<td>0.0842873</td>\n",
       "<td>0.4804992</td>\n",
       "<td>1.1837589</td>\n",
       "<td>0.0873801</td>\n",
       "<td>0.2152698</td>\n",
       "<td>0.0480536</td>\n",
       "<td>0.9470071</td>\n",
       "<td>-51.9500787</td>\n",
       "<td>18.3758895</td></tr>\n",
       "<tr><td></td>\n",
       "<td>15</td>\n",
       "<td>0.8999924</td>\n",
       "<td>0.0587195</td>\n",
       "<td>0.3792670</td>\n",
       "<td>1.0943770</td>\n",
       "<td>0.0689708</td>\n",
       "<td>0.1990155</td>\n",
       "<td>0.0379238</td>\n",
       "<td>0.9849309</td>\n",
       "<td>-62.0732955</td>\n",
       "<td>9.4376960</td></tr>\n",
       "<tr><td></td>\n",
       "<td>16</td>\n",
       "<td>1.0</td>\n",
       "<td>0.0096937</td>\n",
       "<td>0.1506792</td>\n",
       "<td>1.0</td>\n",
       "<td>0.0274014</td>\n",
       "<td>0.1818528</td>\n",
       "<td>0.0150691</td>\n",
       "<td>1.0</td>\n",
       "<td>-84.9320804</td>\n",
       "<td>0.0</td></tr></table></div>"
      ],
      "text/plain": [
       "    group    cumulative_data_fraction    lower_threshold    lift      cumulative_lift    response_rate    cumulative_response_rate    capture_rate    cumulative_capture_rate    gain      cumulative_gain\n",
       "--  -------  --------------------------  -----------------  --------  -----------------  ---------------  --------------------------  --------------  -------------------------  --------  -----------------\n",
       "    1        0.0100023                   0.527574           3.4902    3.4902             0.634703         0.634703                    0.03491         0.03491                    249.02    249.02\n",
       "    2        0.0200046                   0.484513           2.92106   3.20563            0.531202         0.582953                    0.0292172       0.0641272                  192.106   220.563\n",
       "    3        0.0300069                   0.453283           2.63648   3.01592            0.479452         0.548453                    0.0263709       0.0904981                  163.648   201.592\n",
       "    4        0.0400091                   0.430836           2.66996   2.92943            0.48554          0.532725                    0.0267057       0.117204                   166.996   192.943\n",
       "    5        0.0500114                   0.411186           2.65322   2.87419            0.482496         0.522679                    0.0265383       0.143742                   165.322   187.419\n",
       "    6        0.100008                    0.345791           2.21197   2.54313            0.402253         0.462475                    0.11059         0.254332                   121.197   154.313\n",
       "    7        0.150004                    0.301784           1.83689   2.30774            0.334044         0.419669                    0.0918376       0.34617                    83.6892   130.774\n",
       "    8        0.2                         0.269779           1.55893   2.12055            0.283496         0.385628                    0.0779406       0.424111                   55.893    112.055\n",
       "    9        0.300008                    0.221276           1.34021   1.86042            0.243721         0.338323                    0.134031        0.558141                   34.0208   86.0424\n",
       "    10       0.4                         0.184791           1.08506   1.6666             0.19732          0.303075                    0.108497        0.666639                   8.50554   66.6597\n",
       "    11       0.500008                    0.155373           0.929188  1.51911            0.168975         0.276254                    0.0929259       0.759565                   -7.08116  51.9106\n",
       "    12       0.6                         0.130144           0.761046  1.39277            0.138398         0.25328                     0.0760988       0.835663                   -23.8954  39.2772\n",
       "    13       0.699992                    0.107097           0.632949  1.28423            0.115104         0.233541                    0.0632901       0.898954                   -36.7051  28.4233\n",
       "    14       0.8                         0.0842873          0.480499  1.18376            0.0873801        0.21527                     0.0480536       0.947007                   -51.9501  18.3759\n",
       "    15       0.899992                    0.0587195          0.379267  1.09438            0.0689708        0.199015                    0.0379238       0.984931                   -62.0733  9.4377\n",
       "    16       1                           0.0096937          0.150679  1                  0.0274014        0.181853                    0.0150691       1                          -84.9321  0"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "ModelMetricsBinomial: gbm\n",
      "** Reported on validation data. **\n",
      "\n",
      "MSE: 0.137350073063837\n",
      "RMSE: 0.37060770777715485\n",
      "LogLoss: 0.43535916538419495\n",
      "Mean Per-Class Error: 0.3514922046634934\n",
      "AUC: 0.7038515226184847\n",
      "Gini: 0.40770304523696943\n",
      "Confusion Matrix (Act/Pred) for max f1 @ threshold = 0.19924203066741691: \n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div style=\"overflow:auto\"><table style=\"width:50%\"><tr><td><b></b></td>\n",
       "<td><b>0</b></td>\n",
       "<td><b>1</b></td>\n",
       "<td><b>Error</b></td>\n",
       "<td><b>Rate</b></td></tr>\n",
       "<tr><td>0</td>\n",
       "<td>27946.0</td>\n",
       "<td>12240.0</td>\n",
       "<td>0.3046</td>\n",
       "<td> (12240.0/40186.0)</td></tr>\n",
       "<tr><td>1</td>\n",
       "<td>3581.0</td>\n",
       "<td>5379.0</td>\n",
       "<td>0.3997</td>\n",
       "<td> (3581.0/8960.0)</td></tr>\n",
       "<tr><td>Total</td>\n",
       "<td>31527.0</td>\n",
       "<td>17619.0</td>\n",
       "<td>0.3219</td>\n",
       "<td> (15821.0/49146.0)</td></tr></table></div>"
      ],
      "text/plain": [
       "       0      1      Error    Rate\n",
       "-----  -----  -----  -------  -----------------\n",
       "0      27946  12240  0.3046   (12240.0/40186.0)\n",
       "1      3581   5379   0.3997   (3581.0/8960.0)\n",
       "Total  31527  17619  0.3219   (15821.0/49146.0)"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Maximum Metrics: Maximum metrics at their respective thresholds\n",
      "\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div style=\"overflow:auto\"><table style=\"width:50%\"><tr><td><b>metric</b></td>\n",
       "<td><b>threshold</b></td>\n",
       "<td><b>value</b></td>\n",
       "<td><b>idx</b></td></tr>\n",
       "<tr><td>max f1</td>\n",
       "<td>0.1992420</td>\n",
       "<td>0.4047556</td>\n",
       "<td>224.0</td></tr>\n",
       "<tr><td>max f2</td>\n",
       "<td>0.1097674</td>\n",
       "<td>0.5661149</td>\n",
       "<td>311.0</td></tr>\n",
       "<tr><td>max f0point5</td>\n",
       "<td>0.2934679</td>\n",
       "<td>0.3636475</td>\n",
       "<td>152.0</td></tr>\n",
       "<tr><td>max accuracy</td>\n",
       "<td>0.5987741</td>\n",
       "<td>0.8179506</td>\n",
       "<td>20.0</td></tr>\n",
       "<tr><td>max precision</td>\n",
       "<td>0.6706534</td>\n",
       "<td>0.6363636</td>\n",
       "<td>9.0</td></tr>\n",
       "<tr><td>max recall</td>\n",
       "<td>0.0207835</td>\n",
       "<td>1.0</td>\n",
       "<td>396.0</td></tr>\n",
       "<tr><td>max specificity</td>\n",
       "<td>0.8182443</td>\n",
       "<td>0.9999502</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max absolute_mcc</td>\n",
       "<td>0.2139812</td>\n",
       "<td>0.2381822</td>\n",
       "<td>213.0</td></tr>\n",
       "<tr><td>max min_per_class_accuracy</td>\n",
       "<td>0.1825940</td>\n",
       "<td>0.6461454</td>\n",
       "<td>239.0</td></tr>\n",
       "<tr><td>max mean_per_class_accuracy</td>\n",
       "<td>0.1725616</td>\n",
       "<td>0.6485078</td>\n",
       "<td>249.0</td></tr></table></div>"
      ],
      "text/plain": [
       "metric                       threshold    value     idx\n",
       "---------------------------  -----------  --------  -----\n",
       "max f1                       0.199242     0.404756  224\n",
       "max f2                       0.109767     0.566115  311\n",
       "max f0point5                 0.293468     0.363648  152\n",
       "max accuracy                 0.598774     0.817951  20\n",
       "max precision                0.670653     0.636364  9\n",
       "max recall                   0.0207835    1         396\n",
       "max specificity              0.818244     0.99995   0\n",
       "max absolute_mcc             0.213981     0.238182  213\n",
       "max min_per_class_accuracy   0.182594     0.646145  239\n",
       "max mean_per_class_accuracy  0.172562     0.648508  249"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Gains/Lift Table: Avg response rate: 18.23 %\n",
      "\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div style=\"overflow:auto\"><table style=\"width:50%\"><tr><td><b></b></td>\n",
       "<td><b>group</b></td>\n",
       "<td><b>cumulative_data_fraction</b></td>\n",
       "<td><b>lower_threshold</b></td>\n",
       "<td><b>lift</b></td>\n",
       "<td><b>cumulative_lift</b></td>\n",
       "<td><b>response_rate</b></td>\n",
       "<td><b>cumulative_response_rate</b></td>\n",
       "<td><b>capture_rate</b></td>\n",
       "<td><b>cumulative_capture_rate</b></td>\n",
       "<td><b>gain</b></td>\n",
       "<td><b>cumulative_gain</b></td></tr>\n",
       "<tr><td></td>\n",
       "<td>1</td>\n",
       "<td>0.0100110</td>\n",
       "<td>0.5246919</td>\n",
       "<td>2.4972561</td>\n",
       "<td>2.4972561</td>\n",
       "<td>0.4552846</td>\n",
       "<td>0.4552846</td>\n",
       "<td>0.025</td>\n",
       "<td>0.025</td>\n",
       "<td>149.7256098</td>\n",
       "<td>149.7256098</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2</td>\n",
       "<td>0.0200016</td>\n",
       "<td>0.4787756</td>\n",
       "<td>2.5581980</td>\n",
       "<td>2.5276961</td>\n",
       "<td>0.4663951</td>\n",
       "<td>0.4608342</td>\n",
       "<td>0.0255580</td>\n",
       "<td>0.0505580</td>\n",
       "<td>155.8198011</td>\n",
       "<td>152.7696056</td></tr>\n",
       "<tr><td></td>\n",
       "<td>3</td>\n",
       "<td>0.0300126</td>\n",
       "<td>0.4493471</td>\n",
       "<td>2.5864438</td>\n",
       "<td>2.5472919</td>\n",
       "<td>0.4715447</td>\n",
       "<td>0.4644068</td>\n",
       "<td>0.0258929</td>\n",
       "<td>0.0764509</td>\n",
       "<td>158.6443815</td>\n",
       "<td>154.7291919</td></tr>\n",
       "<tr><td></td>\n",
       "<td>4</td>\n",
       "<td>0.0400033</td>\n",
       "<td>0.4265645</td>\n",
       "<td>2.2789188</td>\n",
       "<td>2.4802669</td>\n",
       "<td>0.4154786</td>\n",
       "<td>0.4521872</td>\n",
       "<td>0.0227679</td>\n",
       "<td>0.0992187</td>\n",
       "<td>127.8918752</td>\n",
       "<td>148.0266881</td></tr>\n",
       "<tr><td></td>\n",
       "<td>5</td>\n",
       "<td>0.0500142</td>\n",
       "<td>0.4088086</td>\n",
       "<td>2.1962476</td>\n",
       "<td>2.4234168</td>\n",
       "<td>0.4004065</td>\n",
       "<td>0.4418226</td>\n",
       "<td>0.0219866</td>\n",
       "<td>0.1212054</td>\n",
       "<td>119.6247550</td>\n",
       "<td>142.3416795</td></tr>\n",
       "<tr><td></td>\n",
       "<td>6</td>\n",
       "<td>0.1000081</td>\n",
       "<td>0.3417590</td>\n",
       "<td>2.0448925</td>\n",
       "<td>2.2341932</td>\n",
       "<td>0.3728124</td>\n",
       "<td>0.4073245</td>\n",
       "<td>0.1022321</td>\n",
       "<td>0.2234375</td>\n",
       "<td>104.4892508</td>\n",
       "<td>123.4193159</td></tr>\n",
       "<tr><td></td>\n",
       "<td>7</td>\n",
       "<td>0.1500020</td>\n",
       "<td>0.3008643</td>\n",
       "<td>1.7435164</td>\n",
       "<td>2.0706564</td>\n",
       "<td>0.3178673</td>\n",
       "<td>0.3775095</td>\n",
       "<td>0.0871652</td>\n",
       "<td>0.3106027</td>\n",
       "<td>74.3516429</td>\n",
       "<td>107.0656435</td></tr>\n",
       "<tr><td></td>\n",
       "<td>8</td>\n",
       "<td>0.2000163</td>\n",
       "<td>0.2694811</td>\n",
       "<td>1.6111482</td>\n",
       "<td>1.9557560</td>\n",
       "<td>0.2937347</td>\n",
       "<td>0.3565615</td>\n",
       "<td>0.0805804</td>\n",
       "<td>0.3911830</td>\n",
       "<td>61.1148182</td>\n",
       "<td>95.5755999</td></tr>\n",
       "<tr><td></td>\n",
       "<td>9</td>\n",
       "<td>0.3000041</td>\n",
       "<td>0.2213539</td>\n",
       "<td>1.3963758</td>\n",
       "<td>1.7693212</td>\n",
       "<td>0.2545788</td>\n",
       "<td>0.3225719</td>\n",
       "<td>0.1396205</td>\n",
       "<td>0.5308036</td>\n",
       "<td>39.6375834</td>\n",
       "<td>76.9321237</td></tr>\n",
       "<tr><td></td>\n",
       "<td>10</td>\n",
       "<td>0.4000122</td>\n",
       "<td>0.1844368</td>\n",
       "<td>1.1170966</td>\n",
       "<td>1.6062568</td>\n",
       "<td>0.2036623</td>\n",
       "<td>0.2928430</td>\n",
       "<td>0.1117188</td>\n",
       "<td>0.6425223</td>\n",
       "<td>11.7096579</td>\n",
       "<td>60.6256779</td></tr>\n",
       "<tr><td></td>\n",
       "<td>11</td>\n",
       "<td>0.5</td>\n",
       "<td>0.1557807</td>\n",
       "<td>0.9465441</td>\n",
       "<td>1.4743304</td>\n",
       "<td>0.1725682</td>\n",
       "<td>0.2687909</td>\n",
       "<td>0.0946429</td>\n",
       "<td>0.7371652</td>\n",
       "<td>-5.3455870</td>\n",
       "<td>47.4330357</td></tr>\n",
       "<tr><td></td>\n",
       "<td>12</td>\n",
       "<td>0.6000081</td>\n",
       "<td>0.1302401</td>\n",
       "<td>0.8135499</td>\n",
       "<td>1.3641928</td>\n",
       "<td>0.1483215</td>\n",
       "<td>0.2487113</td>\n",
       "<td>0.0813616</td>\n",
       "<td>0.8185268</td>\n",
       "<td>-18.6450144</td>\n",
       "<td>36.4192804</td></tr>\n",
       "<tr><td></td>\n",
       "<td>13</td>\n",
       "<td>0.6999959</td>\n",
       "<td>0.1067658</td>\n",
       "<td>0.6875839</td>\n",
       "<td>1.2675456</td>\n",
       "<td>0.1253561</td>\n",
       "<td>0.2310912</td>\n",
       "<td>0.06875</td>\n",
       "<td>0.8872768</td>\n",
       "<td>-31.2416056</td>\n",
       "<td>26.7545634</td></tr>\n",
       "<tr><td></td>\n",
       "<td>14</td>\n",
       "<td>0.8000041</td>\n",
       "<td>0.0835078</td>\n",
       "<td>0.5122351</td>\n",
       "<td>1.1731246</td>\n",
       "<td>0.0933876</td>\n",
       "<td>0.2138769</td>\n",
       "<td>0.0512277</td>\n",
       "<td>0.9385045</td>\n",
       "<td>-48.7764905</td>\n",
       "<td>17.3124613</td></tr>\n",
       "<tr><td></td>\n",
       "<td>15</td>\n",
       "<td>0.8999919</td>\n",
       "<td>0.0583687</td>\n",
       "<td>0.4018348</td>\n",
       "<td>1.0874354</td>\n",
       "<td>0.0732601</td>\n",
       "<td>0.1982546</td>\n",
       "<td>0.0401786</td>\n",
       "<td>0.9786830</td>\n",
       "<td>-59.8165228</td>\n",
       "<td>8.7435429</td></tr>\n",
       "<tr><td></td>\n",
       "<td>16</td>\n",
       "<td>1.0</td>\n",
       "<td>0.0097817</td>\n",
       "<td>0.2131523</td>\n",
       "<td>1.0</td>\n",
       "<td>0.0388606</td>\n",
       "<td>0.1823139</td>\n",
       "<td>0.0213170</td>\n",
       "<td>1.0</td>\n",
       "<td>-78.6847706</td>\n",
       "<td>0.0</td></tr></table></div>"
      ],
      "text/plain": [
       "    group    cumulative_data_fraction    lower_threshold    lift      cumulative_lift    response_rate    cumulative_response_rate    capture_rate    cumulative_capture_rate    gain      cumulative_gain\n",
       "--  -------  --------------------------  -----------------  --------  -----------------  ---------------  --------------------------  --------------  -------------------------  --------  -----------------\n",
       "    1        0.010011                    0.524692           2.49726   2.49726            0.455285         0.455285                    0.025           0.025                      149.726   149.726\n",
       "    2        0.0200016                   0.478776           2.5582    2.5277             0.466395         0.460834                    0.025558        0.050558                   155.82    152.77\n",
       "    3        0.0300126                   0.449347           2.58644   2.54729            0.471545         0.464407                    0.0258929       0.0764509                  158.644   154.729\n",
       "    4        0.0400033                   0.426564           2.27892   2.48027            0.415479         0.452187                    0.0227679       0.0992187                  127.892   148.027\n",
       "    5        0.0500142                   0.408809           2.19625   2.42342            0.400407         0.441823                    0.0219866       0.121205                   119.625   142.342\n",
       "    6        0.100008                    0.341759           2.04489   2.23419            0.372812         0.407325                    0.102232        0.223438                   104.489   123.419\n",
       "    7        0.150002                    0.300864           1.74352   2.07066            0.317867         0.377509                    0.0871652       0.310603                   74.3516   107.066\n",
       "    8        0.200016                    0.269481           1.61115   1.95576            0.293735         0.356562                    0.0805804       0.391183                   61.1148   95.5756\n",
       "    9        0.300004                    0.221354           1.39638   1.76932            0.254579         0.322572                    0.139621        0.530804                   39.6376   76.9321\n",
       "    10       0.400012                    0.184437           1.1171    1.60626            0.203662         0.292843                    0.111719        0.642522                   11.7097   60.6257\n",
       "    11       0.5                         0.155781           0.946544  1.47433            0.172568         0.268791                    0.0946429       0.737165                   -5.34559  47.433\n",
       "    12       0.600008                    0.13024            0.81355   1.36419            0.148321         0.248711                    0.0813616       0.818527                   -18.645   36.4193\n",
       "    13       0.699996                    0.106766           0.687584  1.26755            0.125356         0.231091                    0.06875         0.887277                   -31.2416  26.7546\n",
       "    14       0.800004                    0.0835078          0.512235  1.17312            0.0933876        0.213877                    0.0512277       0.938504                   -48.7765  17.3125\n",
       "    15       0.899992                    0.0583687          0.401835  1.08744            0.0732601        0.198255                    0.0401786       0.978683                   -59.8165  8.74354\n",
       "    16       1                           0.00978174         0.213152  1                  0.0388606        0.182314                    0.021317        1                          -78.6848  0"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Scoring History: \n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div style=\"overflow:auto\"><table style=\"width:50%\"><tr><td><b></b></td>\n",
       "<td><b>timestamp</b></td>\n",
       "<td><b>duration</b></td>\n",
       "<td><b>number_of_trees</b></td>\n",
       "<td><b>training_rmse</b></td>\n",
       "<td><b>training_logloss</b></td>\n",
       "<td><b>training_auc</b></td>\n",
       "<td><b>training_lift</b></td>\n",
       "<td><b>training_classification_error</b></td>\n",
       "<td><b>validation_rmse</b></td>\n",
       "<td><b>validation_logloss</b></td>\n",
       "<td><b>validation_auc</b></td>\n",
       "<td><b>validation_lift</b></td>\n",
       "<td><b>validation_classification_error</b></td></tr>\n",
       "<tr><td></td>\n",
       "<td>2017-07-06 17:02:44</td>\n",
       "<td> 3 min 13.668 sec</td>\n",
       "<td>0.0</td>\n",
       "<td>0.3857231</td>\n",
       "<td>0.4741914</td>\n",
       "<td>0.5</td>\n",
       "<td>1.0</td>\n",
       "<td>0.8181472</td>\n",
       "<td>0.3861033</td>\n",
       "<td>0.4748848</td>\n",
       "<td>0.5</td>\n",
       "<td>1.0</td>\n",
       "<td>0.8176861</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2017-07-06 17:02:44</td>\n",
       "<td> 3 min 13.704 sec</td>\n",
       "<td>1.0</td>\n",
       "<td>0.3839030</td>\n",
       "<td>0.4695181</td>\n",
       "<td>0.6579641</td>\n",
       "<td>1.7274827</td>\n",
       "<td>0.3387531</td>\n",
       "<td>0.3843575</td>\n",
       "<td>0.4703953</td>\n",
       "<td>0.6522118</td>\n",
       "<td>1.6793986</td>\n",
       "<td>0.3426322</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2017-07-06 17:02:44</td>\n",
       "<td> 3 min 13.744 sec</td>\n",
       "<td>2.0</td>\n",
       "<td>0.3824515</td>\n",
       "<td>0.4659005</td>\n",
       "<td>0.6720541</td>\n",
       "<td>2.1212783</td>\n",
       "<td>0.3915658</td>\n",
       "<td>0.3829791</td>\n",
       "<td>0.4669506</td>\n",
       "<td>0.6662733</td>\n",
       "<td>1.9709095</td>\n",
       "<td>0.3928906</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2017-07-06 17:02:44</td>\n",
       "<td> 3 min 13.794 sec</td>\n",
       "<td>3.0</td>\n",
       "<td>0.3814437</td>\n",
       "<td>0.4634175</td>\n",
       "<td>0.6754474</td>\n",
       "<td>2.3743959</td>\n",
       "<td>0.3501256</td>\n",
       "<td>0.3820071</td>\n",
       "<td>0.4645476</td>\n",
       "<td>0.6699532</td>\n",
       "<td>2.2953480</td>\n",
       "<td>0.3550848</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2017-07-06 17:02:44</td>\n",
       "<td> 3 min 13.848 sec</td>\n",
       "<td>4.0</td>\n",
       "<td>0.3802737</td>\n",
       "<td>0.4604705</td>\n",
       "<td>0.6803222</td>\n",
       "<td>2.3743959</td>\n",
       "<td>0.3680292</td>\n",
       "<td>0.3808862</td>\n",
       "<td>0.4617129</td>\n",
       "<td>0.6745336</td>\n",
       "<td>2.2953480</td>\n",
       "<td>0.4020063</td></tr>\n",
       "<tr><td>---</td>\n",
       "<td>---</td>\n",
       "<td>---</td>\n",
       "<td>---</td>\n",
       "<td>---</td>\n",
       "<td>---</td>\n",
       "<td>---</td>\n",
       "<td>---</td>\n",
       "<td>---</td>\n",
       "<td>---</td>\n",
       "<td>---</td>\n",
       "<td>---</td>\n",
       "<td>---</td>\n",
       "<td>---</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2017-07-06 17:02:47</td>\n",
       "<td> 3 min 16.418 sec</td>\n",
       "<td>20.0</td>\n",
       "<td>0.3729513</td>\n",
       "<td>0.4424300</td>\n",
       "<td>0.6937648</td>\n",
       "<td>2.8815838</td>\n",
       "<td>0.3425744</td>\n",
       "<td>0.3742133</td>\n",
       "<td>0.4449977</td>\n",
       "<td>0.6872783</td>\n",
       "<td>2.5332104</td>\n",
       "<td>0.3843853</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2017-07-06 17:02:47</td>\n",
       "<td> 3 min 16.722 sec</td>\n",
       "<td>21.0</td>\n",
       "<td>0.3727846</td>\n",
       "<td>0.4420412</td>\n",
       "<td>0.6941957</td>\n",
       "<td>2.8915470</td>\n",
       "<td>0.3464261</td>\n",
       "<td>0.3740674</td>\n",
       "<td>0.4446485</td>\n",
       "<td>0.6876986</td>\n",
       "<td>2.5373730</td>\n",
       "<td>0.3873153</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2017-07-06 17:02:47</td>\n",
       "<td> 3 min 17.034 sec</td>\n",
       "<td>22.0</td>\n",
       "<td>0.3726401</td>\n",
       "<td>0.4416442</td>\n",
       "<td>0.6942048</td>\n",
       "<td>2.9131361</td>\n",
       "<td>0.3313999</td>\n",
       "<td>0.3739474</td>\n",
       "<td>0.4443078</td>\n",
       "<td>0.6876354</td>\n",
       "<td>2.4921907</td>\n",
       "<td>0.3874374</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2017-07-06 17:02:48</td>\n",
       "<td> 3 min 17.365 sec</td>\n",
       "<td>23.0</td>\n",
       "<td>0.3724694</td>\n",
       "<td>0.4412595</td>\n",
       "<td>0.6948000</td>\n",
       "<td>2.9043179</td>\n",
       "<td>0.3325112</td>\n",
       "<td>0.3738226</td>\n",
       "<td>0.4440190</td>\n",
       "<td>0.6880199</td>\n",
       "<td>2.5307015</td>\n",
       "<td>0.3909169</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2017-07-06 17:02:50</td>\n",
       "<td> 3 min 19.963 sec</td>\n",
       "<td>350.0</td>\n",
       "<td>0.3648284</td>\n",
       "<td>0.4234491</td>\n",
       "<td>0.7267850</td>\n",
       "<td>3.4902034</td>\n",
       "<td>0.2824998</td>\n",
       "<td>0.3706077</td>\n",
       "<td>0.4353592</td>\n",
       "<td>0.7038515</td>\n",
       "<td>2.4972561</td>\n",
       "<td>0.3219184</td></tr></table></div>"
      ],
      "text/plain": [
       "     timestamp            duration          number_of_trees    training_rmse        training_logloss     training_auc        training_lift       training_classification_error    validation_rmse      validation_logloss    validation_auc      validation_lift     validation_classification_error\n",
       "---  -------------------  ----------------  -----------------  -------------------  -------------------  ------------------  ------------------  -------------------------------  -------------------  --------------------  ------------------  ------------------  ---------------------------------\n",
       "     2017-07-06 17:02:44  3 min 13.668 sec  0.0                0.38572314918565614  0.4741913507137309   0.5                 1.0                 0.8181472177818375               0.38610331306206835  0.4748848330622531    0.5                 1.0                 0.8176860782159281\n",
       "     2017-07-06 17:02:44  3 min 13.704 sec  1.0                0.38390301840464697  0.4695180693734886   0.6579640784746738  1.7274827132246144  0.33875313998629825              0.3843575454734482   0.4703953467668367    0.6522117841212469  1.6793986036204585  0.3426321572457575\n",
       "     2017-07-06 17:02:44  3 min 13.744 sec  2.0                0.3824515206242979   0.4659004819518625   0.6720541394055342  2.1212782808495776  0.39156580650072315              0.3829791476593432   0.4669505924437416    0.6662732940265267  1.970909478557505   0.39289057095185775\n",
       "     2017-07-06 17:02:44  3 min 13.794 sec  3.0                0.38144373620663335  0.46341747086200774  0.6754474141265566  2.3743958938884013  0.35012559945192967              0.38200707798925637  0.46454764788969555   0.6699532344797584  2.2953479991394152  0.35508484922475886\n",
       "     2017-07-06 17:02:44  3 min 13.848 sec  4.0                0.3802736757886236   0.4604705329333941   0.6803222334783088  2.3743958938884013  0.36802923041790364              0.3808861939673367   0.46171287765675756   0.6745336487231695  2.2953479991394152  0.40200626704106135\n",
       "---  ---                  ---               ---                ---                  ---                  ---                 ---                 ---                              ---                  ---                   ---                 ---                 ---\n",
       "     2017-07-06 17:02:47  3 min 16.418 sec  20.0               0.3729512897315818   0.44242995019378595  0.6937648364456681  2.88158376040335    0.34257440816015833              0.374213332115361    0.44499766320496087   0.6872783118210144  2.533210441710931   0.3843853009400562\n",
       "     2017-07-06 17:02:47  3 min 16.722 sec  21.0               0.3727846487012486   0.4420411598287236   0.6941957042286762  2.891546996683126   0.34642612468600137              0.374067398571832    0.44464851365850294   0.6876986118899795  2.5373730139201354  0.3873153461115859\n",
       "     2017-07-06 17:02:47  3 min 17.034 sec  22.0               0.37264012641745287  0.441644202105135    0.6942048408200158  2.9131360999820606  0.3313998629824161               0.3739473967806829   0.44430781498045213   0.6876353624729828  2.492190669371197   0.3874374313270663\n",
       "     2017-07-06 17:02:48  3 min 17.365 sec  23.0               0.37246940411536295  0.4412594553914065   0.6947999951084575  2.9043179259582064  0.3325112278297937               0.3738226173795527   0.4440190203514056    0.6880198955437572  2.530701491724739   0.39091685996825787\n",
       "     2017-07-06 17:02:50  3 min 19.963 sec  350.0              0.36482843147255     0.4234491377496304   0.7267849815001551  3.490203386526144   0.2824998096978001               0.37060770777715485  0.43535916538419495   0.7038515226184847  2.497256097560976   0.3219183656859154"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "See the whole table with table.as_data_frame()\n",
      "Variable Importances: \n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div style=\"overflow:auto\"><table style=\"width:50%\"><tr><td><b>variable</b></td>\n",
       "<td><b>relative_importance</b></td>\n",
       "<td><b>scaled_importance</b></td>\n",
       "<td><b>percentage</b></td></tr>\n",
       "<tr><td>int_rate</td>\n",
       "<td>2268.7045898</td>\n",
       "<td>1.0</td>\n",
       "<td>0.4653831</td></tr>\n",
       "<tr><td>addr_state</td>\n",
       "<td>758.4807739</td>\n",
       "<td>0.3343233</td>\n",
       "<td>0.1555884</td></tr>\n",
       "<tr><td>annual_inc</td>\n",
       "<td>430.6903687</td>\n",
       "<td>0.1898398</td>\n",
       "<td>0.0883482</td></tr>\n",
       "<tr><td>term</td>\n",
       "<td>358.3098145</td>\n",
       "<td>0.1579359</td>\n",
       "<td>0.0735007</td></tr>\n",
       "<tr><td>dti</td>\n",
       "<td>274.9893188</td>\n",
       "<td>0.1212098</td>\n",
       "<td>0.0564090</td></tr>\n",
       "<tr><td>purpose</td>\n",
       "<td>221.0051575</td>\n",
       "<td>0.0974147</td>\n",
       "<td>0.0453352</td></tr>\n",
       "<tr><td>revol_util</td>\n",
       "<td>173.9438629</td>\n",
       "<td>0.0766710</td>\n",
       "<td>0.0356814</td></tr>\n",
       "<tr><td>loan_amnt</td>\n",
       "<td>133.6167908</td>\n",
       "<td>0.0588956</td>\n",
       "<td>0.0274090</td></tr>\n",
       "<tr><td>total_acc</td>\n",
       "<td>76.3822479</td>\n",
       "<td>0.0336678</td>\n",
       "<td>0.0156684</td></tr>\n",
       "<tr><td>emp_length</td>\n",
       "<td>72.7174759</td>\n",
       "<td>0.0320524</td>\n",
       "<td>0.0149167</td></tr>\n",
       "<tr><td>home_ownership</td>\n",
       "<td>52.8385696</td>\n",
       "<td>0.0232902</td>\n",
       "<td>0.0108389</td></tr>\n",
       "<tr><td>longest_credit_length</td>\n",
       "<td>31.1078701</td>\n",
       "<td>0.0137117</td>\n",
       "<td>0.0063812</td></tr>\n",
       "<tr><td>delinq_2yrs</td>\n",
       "<td>20.7606449</td>\n",
       "<td>0.0091509</td>\n",
       "<td>0.0042587</td></tr>\n",
       "<tr><td>verification_status</td>\n",
       "<td>1.3708715</td>\n",
       "<td>0.0006043</td>\n",
       "<td>0.0002812</td></tr></table></div>"
      ],
      "text/plain": [
       "variable               relative_importance    scaled_importance    percentage\n",
       "---------------------  ---------------------  -------------------  ------------\n",
       "int_rate               2268.7                 1                    0.465383\n",
       "addr_state             758.481                0.334323             0.155588\n",
       "annual_inc             430.69                 0.18984              0.0883482\n",
       "term                   358.31                 0.157936             0.0735007\n",
       "dti                    274.989                0.12121              0.056409\n",
       "purpose                221.005                0.0974147            0.0453352\n",
       "revol_util             173.944                0.076671             0.0356814\n",
       "loan_amnt              133.617                0.0588956            0.027409\n",
       "total_acc              76.3822                0.0336678            0.0156684\n",
       "emp_length             72.7175                0.0320524            0.0149167\n",
       "home_ownership         52.8386                0.0232902            0.0108389\n",
       "longest_credit_length  31.1079                0.0137117            0.00638121\n",
       "delinq_2yrs            20.7606                0.00915088           0.00425867\n",
       "verification_status    1.37087                0.000604253          0.000281209"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": []
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# show grid search results\n",
    "gsearch.show()\n",
    "\n",
    "# select best model\n",
    "gbm_model = gsearch.get_grid()[0]\n",
    "\n",
    "# print model information\n",
    "gbm_model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.7267849815001551\n",
      "0.7038515226184847\n",
      "0.7051020666726335\n"
     ]
    }
   ],
   "source": [
    "# measure gbm AUC\n",
    "print(gbm_model.auc(train=True))\n",
    "print(gbm_model.auc(valid=True))\n",
    "print(gbm_model.model_performance(test_data=test).auc())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABLMAAANLCAYAAAC65N+YAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x11515a438>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# analyze variable importance\n",
    "gbm_model.varimp_plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "PartialDependencePlot progress: |█████████████████████████████████████████| 100%\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAsQAAAPxCAYAAAD9lbspAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x11abed320>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# partial dependence plots are a powerful machine learning interpretation tool\n",
    "# to calculate partial dependence across the domain a variable\n",
    "# hold column of interest at constant value\n",
    "# find the mean prediction of the model with this column constant\n",
    "# repeat for multiple values of the variable of interest\n",
    "# h2o has a built-in function for partial dependence as well\n",
    "par_dep_dti1 = gbm_model.partial_plot(data=train, cols=['dti'], server=True, plot=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Are you sure you want to shutdown the H2O instance running at http://127.0.0.1:54321 (Y/N)? y\n",
      "H2O session _sid_95b0 closed.\n"
     ]
    }
   ],
   "source": [
    "# shutdown h2o ... be careful this can erase your work\n",
    "h2o.cluster().shutdown(prompt=True)"
   ]
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "kernelspec": {
   "display_name": "Python [conda root]",
   "language": "python",
   "name": "conda-root-py"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.5.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}