{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## License \n",
    "\n",
    "Copyright 2017 - 2020 Patrick Hall and the H2O.ai team\n",
    "\n",
    "Licensed under the Apache License, Version 2.0 (the \"License\");\n",
    "you may not use this file except in compliance with the License.\n",
    "You may obtain a copy of the License at\n",
    "\n",
    "    http://www.apache.org/licenses/LICENSE-2.0\n",
    "\n",
    "Unless required by applicable law or agreed to in writing, software\n",
    "distributed under the License is distributed on an \"AS IS\" BASIS,\n",
    "WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.\n",
    "See the License for the specific language governing permissions and\n",
    "limitations under the License."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**DISCLAIMER:** This notebook is not legal compliance advice."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Increase Transparency and Accountability in Your Machine Learning Project with Python and H2O\n",
    "#### Explain your complex models with decision tree surrogates, GBM feature importance, and reason codes"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Decision trees and decision tree ensembles are some of the most popular machine learning models used in commercial practice. They can train and make predictions on data containing character values and missing values - both common in large commercial data stores. Single decision trees are easily represented as directed graphs, which can drastically increase their interpretability and transparency. Decision tree ensembles (i.e., random forests and gradient boosting machines (GBMs)), can be used to increase the accuracy and stability of single decision tree models, but are far less intepretable than single trees. These characteristics of decision trees will be leveraged here to increase transparency and accountability in complex, nonlinear, machine learning models.\n",
    "\n",
    "This notebook starts by training a GBM on the UCI credit card default data using the popular open source library, h2o. A single decision tree *surrogate* model will then be trained on the original UCI credit card default data and the predictions from the h2o GBM, to create an approximate flow chart for the GBM's global decision-making processes. A technique known as leave-one-covariate/column-out (LOCO) will then be used to generate local explanations for any row-wise prediction made by the GBM model. Finally, local explanations are ensembled together from multiple similar models to increase explanation stability. \n",
    "\n",
    "**Note**: As of the h2o 3.24 \"Yates\" release, Shapley values are supported in h2o and LOCO is no longer recommended. To see Shapley values for an h2o GBM in action please see: https://github.com/jphall663/interpretable_machine_learning_with_python/blob/master/dia.ipynb."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Python imports\n",
    "In general, NumPy and Pandas will be used for data manipulation purposes and h2o will be used for modeling tasks. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "# imports\n",
    "\n",
    "# h2o Python API with specific classes\n",
    "import h2o                                        \n",
    "from h2o.estimators.gbm import H2OGradientBoostingEstimator       # for GBM\n",
    "from h2o.estimators.random_forest import H2ORandomForestEstimator # for single tree\n",
    "from h2o.backend import H2OLocalServer                            # for plotting local tree in-notebook\n",
    "\n",
    "import numpy as np   # array, vector, matrix calculations\n",
    "import pandas as pd  # DataFrame handling\n",
    "\n",
    "# system packages for calling external graphviz processes\n",
    "import os\n",
    "import re\n",
    "import subprocess\n",
    "\n",
    "# in-notebook display\n",
    "from IPython.display import Image\n",
    "from IPython.display import display\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Start h2o\n",
    "H2o is both a library and a server. The machine learning algorithms in the library take advantage of the multithreaded and distributed architecture provided by the server to train machine learning algorithms extremely efficiently. The API for the library was imported above in cell 1, but the server still needs to be started."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "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: openjdk version \"1.8.0_232\"; OpenJDK Runtime Environment (build 1.8.0_232-8u232-b09-0ubuntu1~16.04.1-b09); OpenJDK 64-Bit Server VM (build 25.232-b09, mixed mode)\n",
      "  Starting server from /home/patrickh/workspace/interpretable_machine_learning_with_python/env_iml/lib/python3.6/site-packages/h2o/backend/bin/h2o.jar\n",
      "  Ice root: /tmp/tmpjpy5faje\n",
      "  JVM stdout: /tmp/tmpjpy5faje/h2o_patrickh_started_from_python.out\n",
      "  JVM stderr: /tmp/tmpjpy5faje/h2o_patrickh_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",
      "Warning: Your H2O cluster version is too old (4 months and 14 days)! Please download and install the latest version from http://h2o.ai/download/\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div style=\"overflow:auto\"><table style=\"width:50%\"><tr><td>H2O cluster uptime:</td>\n",
       "<td>01 secs</td></tr>\n",
       "<tr><td>H2O cluster timezone:</td>\n",
       "<td>America/New_York</td></tr>\n",
       "<tr><td>H2O data parsing timezone:</td>\n",
       "<td>UTC</td></tr>\n",
       "<tr><td>H2O cluster version:</td>\n",
       "<td>3.26.0.3</td></tr>\n",
       "<tr><td>H2O cluster version age:</td>\n",
       "<td>4 months and 14 days !!!</td></tr>\n",
       "<tr><td>H2O cluster name:</td>\n",
       "<td>H2O_from_python_patrickh_6414b6</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>1.778 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>H2O API Extensions:</td>\n",
       "<td>Amazon S3, XGBoost, Algos, AutoML, Core V3, Core V4</td></tr>\n",
       "<tr><td>Python version:</td>\n",
       "<td>3.6.3 final</td></tr></table></div>"
      ],
      "text/plain": [
       "--------------------------  ---------------------------------------------------\n",
       "H2O cluster uptime:         01 secs\n",
       "H2O cluster timezone:       America/New_York\n",
       "H2O data parsing timezone:  UTC\n",
       "H2O cluster version:        3.26.0.3\n",
       "H2O cluster version age:    4 months and 14 days !!!\n",
       "H2O cluster name:           H2O_from_python_patrickh_6414b6\n",
       "H2O cluster total nodes:    1\n",
       "H2O cluster free memory:    1.778 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",
       "H2O API Extensions:         Amazon S3, XGBoost, Algos, AutoML, Core V3, Core V4\n",
       "Python version:             3.6.3 final\n",
       "--------------------------  ---------------------------------------------------"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "h2o.init(max_mem_size='2G')       # start h2o\n",
    "h2o.remove_all()                  # remove any existing data structures from h2o memory"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 1. Download, explore, and prepare UCI credit card default data\n",
    "\n",
    "UCI credit card default data: https://archive.ics.uci.edu/ml/datasets/default+of+credit+card+clients\n",
    "\n",
    "The UCI credit card default data contains demographic and payment information about credit card customers in Taiwan in the year 2005. The data set contains 23 input variables: \n",
    "\n",
    "* **`LIMIT_BAL`**: Amount of given credit (NT dollar)\n",
    "* **`SEX`**: 1 = male; 2 = female\n",
    "* **`EDUCATION`**: 1 = graduate school; 2 = university; 3 = high school; 4 = others \n",
    "* **`MARRIAGE`**: 1 = married; 2 = single; 3 = others\n",
    "* **`AGE`**: Age in years \n",
    "* **`PAY_0`, `PAY_2` - `PAY_6`**: History of past payment; `PAY_0` = the repayment status in September, 2005; `PAY_2` = the repayment status in August, 2005; ...; `PAY_6` = the repayment status in April, 2005. The measurement scale for the repayment status is: -1 = pay duly; 1 = payment delay for one month; 2 = payment delay for two months; ...; 8 = payment delay for eight months; 9 = payment delay for nine months and above. \n",
    "* **`BILL_AMT1` - `BILL_AMT6`**: Amount of bill statement (NT dollar). `BILL_AMNT1` = amount of bill statement in September, 2005; `BILL_AMT2` = amount of bill statement in August, 2005; ...; `BILL_AMT6` = amount of bill statement in April, 2005. \n",
    "* **`PAY_AMT1` - `PAY_AMT6`**: Amount of previous payment (NT dollar). `PAY_AMT1` = amount paid in September, 2005; `PAY_AMT2` = amount paid in August, 2005; ...; `PAY_AMT6` = amount paid in April, 2005. \n",
    "\n",
    "These 23 input variables are used to predict the target variable, whether or not a customer defaulted on their credit card bill in late 2005.\n",
    "\n",
    "Because h2o accepts both numeric and character inputs, some variables will be recoded into more transparent character values."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Import data and clean\n",
    "The credit card default data is available as an `.xls` file. Pandas reads `.xls` files automatically, so it's used to load the credit card default data and give the prediction target a shorter name: `DEFAULT_NEXT_MONTH`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# import XLS file\n",
    "path = 'default_of_credit_card_clients.xls'\n",
    "data = pd.read_excel(path,\n",
    "                     skiprows=1)\n",
    "\n",
    "# remove spaces from target column name \n",
    "data = data.rename(columns={'default payment next month': 'DEFAULT_NEXT_MONTH'}) "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Assign modeling roles\n",
    "The shorthand name `y` is assigned to the prediction target. `X` is assigned to all other input variables in the credit card default data except the row indentifier, `ID`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "y = DEFAULT_NEXT_MONTH\n",
      "X = ['LIMIT_BAL', 'SEX', 'EDUCATION', 'MARRIAGE', 'AGE', 'PAY_0', 'PAY_2', 'PAY_3', 'PAY_4', 'PAY_5', 'PAY_6', 'BILL_AMT1', 'BILL_AMT2', 'BILL_AMT3', 'BILL_AMT4', 'BILL_AMT5', 'BILL_AMT6', 'PAY_AMT1', 'PAY_AMT2', 'PAY_AMT3', 'PAY_AMT4', 'PAY_AMT5', 'PAY_AMT6']\n"
     ]
    }
   ],
   "source": [
    "# assign target and inputs for GBM\n",
    "y = 'DEFAULT_NEXT_MONTH'\n",
    "X = [name for name in data.columns if name not in [y, 'ID']]\n",
    "print('y =', y)\n",
    "print('X =', X)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Helper function for recoding values in the UCI credict card default data\n",
    "This simple function maps longer, more understandable character string values from the UCI credit card default data dictionary to the original integer values of the input variables found in the dataset. These character values can be used directly in h2o decision tree models, and the function returns the original Pandas DataFrame as an h2o object, an H2OFrame. H2o models cannot run on Pandas DataFrames. They require H2OFrames."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Parse progress: |█████████████████████████████████████████████████████████| 100%\n"
     ]
    }
   ],
   "source": [
    "def recode_cc_data(frame):\n",
    "    \n",
    "    \"\"\" Recodes numeric categorical variables into categorical character variables\n",
    "    with more transparent values. \n",
    "    \n",
    "    Args:\n",
    "        frame: Pandas DataFrame version of UCI credit card default data.\n",
    "        \n",
    "    Returns: \n",
    "        H2OFrame with recoded values.\n",
    "        \n",
    "    \"\"\"\n",
    "    \n",
    "    # define recoded values\n",
    "    sex_dict = {1:'male', 2:'female'}\n",
    "    education_dict = {0:'other', 1:'graduate school', 2:'university', 3:'high school', \n",
    "                      4:'other', 5:'other', 6:'other'}\n",
    "    marriage_dict = {0:'other', 1:'married', 2:'single', 3:'divorced'}\n",
    "    pay_dict = {-2:'no consumption', -1:'pay duly', 0:'use of revolving credit', 1:'1 month delay', \n",
    "                2:'2 month delay', 3:'3 month delay', 4:'4 month delay', 5:'5 month delay', 6:'6 month delay', \n",
    "                7:'7 month delay', 8:'8 month delay', 9:'9+ month delay'}\n",
    "    \n",
    "    # recode values using Pandas apply() and anonymous function\n",
    "    frame['SEX'] = frame['SEX'].apply(lambda i: sex_dict[i])\n",
    "    frame['EDUCATION'] = frame['EDUCATION'].apply(lambda i: education_dict[i])    \n",
    "    frame['MARRIAGE'] = frame['MARRIAGE'].apply(lambda i: marriage_dict[i]) \n",
    "    for name in frame.columns:\n",
    "        if name in ['PAY_0', 'PAY_2', 'PAY_3', 'PAY_4', 'PAY_5', 'PAY_6']:\n",
    "            frame[name] = frame[name].apply(lambda i: pay_dict[i])            \n",
    "                \n",
    "    return h2o.H2OFrame(frame)\n",
    "\n",
    "data = recode_cc_data(data)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Ensure target is handled as a categorical variable"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In h2o, a numeric variable can be treated as numeric or categorical. The target variable `DEFAULT_NEXT_MONTH` takes on values of `0` or `1`. To ensure this numeric variable is treated as a categorical variable, the `asfactor()` function is used to explicitly declare that it is a categorical variable. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "data[y] = data[y].asfactor() "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Display descriptive statistics\n",
    "The h2o `describe()` function displays a brief description of the credit card default data. For the categorical input variables `LIMIT_BAL`, `SEX`, `EDUCATION`, `MARRIAGE`, and `PAY_0`-`PAY_6`, the new character values created above in cell 5 are visible. Basic descriptive statistics are displayed for numeric inputs. Also, it's easy to see there are no missing values in this dataset, which will be an important consideration for calculating LOCO values in section 5 and 6."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Rows:30000\n",
      "Cols:24\n",
      "\n",
      "\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<table>\n",
       "<thead>\n",
       "<tr><th>       </th><th>LIMIT_BAL         </th><th>SEX   </th><th>EDUCATION      </th><th>MARRIAGE  </th><th>AGE              </th><th>PAY_0                  </th><th>PAY_2                  </th><th>PAY_3                  </th><th>PAY_4                  </th><th>PAY_5                  </th><th>PAY_6                  </th><th>BILL_AMT1        </th><th>BILL_AMT2        </th><th>BILL_AMT3        </th><th>BILL_AMT4        </th><th>BILL_AMT5        </th><th>BILL_AMT6        </th><th>PAY_AMT1          </th><th>PAY_AMT2          </th><th>PAY_AMT3          </th><th>PAY_AMT4          </th><th>PAY_AMT5          </th><th>PAY_AMT6          </th><th>DEFAULT_NEXT_MONTH  </th></tr>\n",
       "</thead>\n",
       "<tbody>\n",
       "<tr><td>type   </td><td>int               </td><td>enum  </td><td>enum           </td><td>enum      </td><td>int              </td><td>enum                   </td><td>enum                   </td><td>enum                   </td><td>enum                   </td><td>enum                   </td><td>enum                   </td><td>int              </td><td>int              </td><td>int              </td><td>int              </td><td>int              </td><td>int              </td><td>int               </td><td>int               </td><td>int               </td><td>int               </td><td>int               </td><td>int               </td><td>enum                </td></tr>\n",
       "<tr><td>mins   </td><td>10000.0           </td><td>      </td><td>               </td><td>          </td><td>21.0             </td><td>                       </td><td>                       </td><td>                       </td><td>                       </td><td>                       </td><td>                       </td><td>-165580.0        </td><td>-69777.0         </td><td>-157264.0        </td><td>-170000.0        </td><td>-81334.0         </td><td>-339603.0        </td><td>0.0               </td><td>0.0               </td><td>0.0               </td><td>0.0               </td><td>0.0               </td><td>0.0               </td><td>                    </td></tr>\n",
       "<tr><td>mean   </td><td>167484.32266666688</td><td>      </td><td>               </td><td>          </td><td>35.48549999999994</td><td>                       </td><td>                       </td><td>                       </td><td>                       </td><td>                       </td><td>                       </td><td>51223.33090000009</td><td>49179.07516666668</td><td>47013.15479999971</td><td>43262.9489666666 </td><td>40311.40096666653</td><td>38871.76039999991</td><td>5663.580500000014 </td><td>5921.16350000001  </td><td>5225.681500000005 </td><td>4826.076866666661 </td><td>4799.387633333302 </td><td>5215.502566666664 </td><td>                    </td></tr>\n",
       "<tr><td>maxs   </td><td>1000000.0         </td><td>      </td><td>               </td><td>          </td><td>79.0             </td><td>                       </td><td>                       </td><td>                       </td><td>                       </td><td>                       </td><td>                       </td><td>964511.0         </td><td>983931.0         </td><td>1664089.0        </td><td>891586.0         </td><td>927171.0         </td><td>961664.0         </td><td>873552.0          </td><td>1684259.0         </td><td>896040.0          </td><td>621000.0          </td><td>426529.0          </td><td>528666.0          </td><td>                    </td></tr>\n",
       "<tr><td>sigma  </td><td>129747.66156720225</td><td>      </td><td>               </td><td>          </td><td>9.21790406809016 </td><td>                       </td><td>                       </td><td>                       </td><td>                       </td><td>                       </td><td>                       </td><td>73635.86057552959</td><td>71173.76878252836</td><td>69349.38742703681</td><td>64332.85613391641</td><td>60797.1557702648 </td><td>59554.10753674574</td><td>16563.280354025763</td><td>23040.870402057226</td><td>17606.961469803115</td><td>15666.159744031993</td><td>15278.305679144793</td><td>17777.465775435332</td><td>                    </td></tr>\n",
       "<tr><td>zeros  </td><td>0                 </td><td>      </td><td>               </td><td>          </td><td>0                </td><td>                       </td><td>                       </td><td>                       </td><td>                       </td><td>                       </td><td>                       </td><td>2008             </td><td>2506             </td><td>2870             </td><td>3195             </td><td>3506             </td><td>4020             </td><td>5249              </td><td>5396              </td><td>5968              </td><td>6408              </td><td>6703              </td><td>7173              </td><td>                    </td></tr>\n",
       "<tr><td>missing</td><td>0                 </td><td>0     </td><td>0              </td><td>0         </td><td>0                </td><td>0                      </td><td>0                      </td><td>0                      </td><td>0                      </td><td>0                      </td><td>0                      </td><td>0                </td><td>0                </td><td>0                </td><td>0                </td><td>0                </td><td>0                </td><td>0                 </td><td>0                 </td><td>0                 </td><td>0                 </td><td>0                 </td><td>0                 </td><td>0                   </td></tr>\n",
       "<tr><td>0      </td><td>20000.0           </td><td>female</td><td>university     </td><td>married   </td><td>24.0             </td><td>2 month delay          </td><td>2 month delay          </td><td>pay duly               </td><td>pay duly               </td><td>no consumption         </td><td>no consumption         </td><td>3913.0           </td><td>3102.0           </td><td>689.0            </td><td>0.0              </td><td>0.0              </td><td>0.0              </td><td>0.0               </td><td>689.0             </td><td>0.0               </td><td>0.0               </td><td>0.0               </td><td>0.0               </td><td>1                   </td></tr>\n",
       "<tr><td>1      </td><td>120000.0          </td><td>female</td><td>university     </td><td>single    </td><td>26.0             </td><td>pay duly               </td><td>2 month delay          </td><td>use of revolving credit</td><td>use of revolving credit</td><td>use of revolving credit</td><td>2 month delay          </td><td>2682.0           </td><td>1725.0           </td><td>2682.0           </td><td>3272.0           </td><td>3455.0           </td><td>3261.0           </td><td>0.0               </td><td>1000.0            </td><td>1000.0            </td><td>1000.0            </td><td>0.0               </td><td>2000.0            </td><td>1                   </td></tr>\n",
       "<tr><td>2      </td><td>90000.0           </td><td>female</td><td>university     </td><td>single    </td><td>34.0             </td><td>use of revolving credit</td><td>use of revolving credit</td><td>use of revolving credit</td><td>use of revolving credit</td><td>use of revolving credit</td><td>use of revolving credit</td><td>29239.0          </td><td>14027.0          </td><td>13559.0          </td><td>14331.0          </td><td>14948.0          </td><td>15549.0          </td><td>1518.0            </td><td>1500.0            </td><td>1000.0            </td><td>1000.0            </td><td>1000.0            </td><td>5000.0            </td><td>0                   </td></tr>\n",
       "<tr><td>3      </td><td>50000.0           </td><td>female</td><td>university     </td><td>married   </td><td>37.0             </td><td>use of revolving credit</td><td>use of revolving credit</td><td>use of revolving credit</td><td>use of revolving credit</td><td>use of revolving credit</td><td>use of revolving credit</td><td>46990.0          </td><td>48233.0          </td><td>49291.0          </td><td>28314.0          </td><td>28959.0          </td><td>29547.0          </td><td>2000.0            </td><td>2019.0            </td><td>1200.0            </td><td>1100.0            </td><td>1069.0            </td><td>1000.0            </td><td>0                   </td></tr>\n",
       "<tr><td>4      </td><td>50000.0           </td><td>male  </td><td>university     </td><td>married   </td><td>57.0             </td><td>pay duly               </td><td>use of revolving credit</td><td>pay duly               </td><td>use of revolving credit</td><td>use of revolving credit</td><td>use of revolving credit</td><td>8617.0           </td><td>5670.0           </td><td>35835.0          </td><td>20940.0          </td><td>19146.0          </td><td>19131.0          </td><td>2000.0            </td><td>36681.0           </td><td>10000.0           </td><td>9000.0            </td><td>689.0             </td><td>679.0             </td><td>0                   </td></tr>\n",
       "<tr><td>5      </td><td>50000.0           </td><td>male  </td><td>graduate school</td><td>single    </td><td>37.0             </td><td>use of revolving credit</td><td>use of revolving credit</td><td>use of revolving credit</td><td>use of revolving credit</td><td>use of revolving credit</td><td>use of revolving credit</td><td>64400.0          </td><td>57069.0          </td><td>57608.0          </td><td>19394.0          </td><td>19619.0          </td><td>20024.0          </td><td>2500.0            </td><td>1815.0            </td><td>657.0             </td><td>1000.0            </td><td>1000.0            </td><td>800.0             </td><td>0                   </td></tr>\n",
       "<tr><td>6      </td><td>500000.0          </td><td>male  </td><td>graduate school</td><td>single    </td><td>29.0             </td><td>use of revolving credit</td><td>use of revolving credit</td><td>use of revolving credit</td><td>use of revolving credit</td><td>use of revolving credit</td><td>use of revolving credit</td><td>367965.0         </td><td>412023.0         </td><td>445007.0         </td><td>542653.0         </td><td>483003.0         </td><td>473944.0         </td><td>55000.0           </td><td>40000.0           </td><td>38000.0           </td><td>20239.0           </td><td>13750.0           </td><td>13770.0           </td><td>0                   </td></tr>\n",
       "<tr><td>7      </td><td>100000.0          </td><td>female</td><td>university     </td><td>single    </td><td>23.0             </td><td>use of revolving credit</td><td>pay duly               </td><td>pay duly               </td><td>use of revolving credit</td><td>use of revolving credit</td><td>pay duly               </td><td>11876.0          </td><td>380.0            </td><td>601.0            </td><td>221.0            </td><td>-159.0           </td><td>567.0            </td><td>380.0             </td><td>601.0             </td><td>0.0               </td><td>581.0             </td><td>1687.0            </td><td>1542.0            </td><td>0                   </td></tr>\n",
       "<tr><td>8      </td><td>140000.0          </td><td>female</td><td>high school    </td><td>married   </td><td>28.0             </td><td>use of revolving credit</td><td>use of revolving credit</td><td>2 month delay          </td><td>use of revolving credit</td><td>use of revolving credit</td><td>use of revolving credit</td><td>11285.0          </td><td>14096.0          </td><td>12108.0          </td><td>12211.0          </td><td>11793.0          </td><td>3719.0           </td><td>3329.0            </td><td>0.0               </td><td>432.0             </td><td>1000.0            </td><td>1000.0            </td><td>1000.0            </td><td>0                   </td></tr>\n",
       "<tr><td>9      </td><td>20000.0           </td><td>male  </td><td>high school    </td><td>single    </td><td>35.0             </td><td>no consumption         </td><td>no consumption         </td><td>no consumption         </td><td>no consumption         </td><td>pay duly               </td><td>pay duly               </td><td>0.0              </td><td>0.0              </td><td>0.0              </td><td>0.0              </td><td>13007.0          </td><td>13912.0          </td><td>0.0               </td><td>0.0               </td><td>0.0               </td><td>13007.0           </td><td>1122.0            </td><td>0.0               </td><td>0                   </td></tr>\n",
       "</tbody>\n",
       "</table>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "data[X + [y]].describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2. Train an H2O GBM classifier"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Split data into training and test sets for early stopping\n",
    "The credit card default data is split into training and test sets to monitor and prevent overtraining. Reproducibility is also an important factor in creating trustworthy models, and randomly splitting datasets can introduce randomness in model predictions and other results. A random seed is used here to ensure the data split is reproducible."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train data rows = 21060, columns = 25\n",
      "Test data rows = 8940, columns = 25\n"
     ]
    }
   ],
   "source": [
    "# split into training and validation\n",
    "train, test = data.split_frame([0.7], seed=12345)\n",
    "\n",
    "# summarize split\n",
    "print('Train data rows = %d, columns = %d' % (train.shape[0], train.shape[1]))\n",
    "print('Test data rows = %d, columns = %d' % (test.shape[0], test.shape[1]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Train h2o GBM classifier\n",
    "Many tuning parameters must be specified to train a GBM using h2o. Typically a grid search would be performed to identify the best parameters for a given modeling task using the `H2OGridSearch` class. For brevity's sake, a previously-discovered set of good tuning parameters are specified here. Because gradient boosting methods typically resample training data, an additional random seed is also specified for the h2o GBM using the `seed` parameter to create reproducible predictions, error rates, and variable importance values. To avoid overfitting, the `stopping_rounds` parameter is used to stop the training process after the test error fails to decrease for 5 iterations.\n",
    "\n",
    "The `balance_classes` parameter ensures the positive and negative classes of the target variable are seen by the model in equal proportions during training. This can be very important for the LOCO calculations in section 5 and 6 for unbalanced data. From experiments across several data sets, explanations for rows with a majority class label for the target variable (e.g., 0) generated by LOCO are more likely to match those generated by another popular explanatory technique, LIME, when the target class is rebalanced during training. `balance_classes` is commented below because the row explained in this notebook has a minority class label (e.g., 1)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "gbm Model Build progress: |███████████████████████████████████████████████| 100%\n",
      "GBM Test AUC = 0.78\n"
     ]
    }
   ],
   "source": [
    "# initialize GBM model\n",
    "model = H2OGradientBoostingEstimator(ntrees=150,            # maximum 150 trees in GBM\n",
    "                                     max_depth=4,           # trees can have maximum depth of 4\n",
    "                                     sample_rate=0.9,       # use 90% of rows in each iteration (tree)\n",
    "                                     col_sample_rate=0.9,   # use 90% of variables in each iteration (tree)\n",
    "                                     #balance_classes=True, # sample to balance 0/1 distribution of target - can help LOCO\n",
    "                                     stopping_rounds=5,     # stop if validation error does not decrease for 5 iterations (trees)\n",
    "                                     score_tree_interval=1, # for reproducibility, set higher for bigger data\n",
    "                                     seed=12345)            # for reproducibility\n",
    "\n",
    "# train a GBM model\n",
    "model.train(y=y, x=X, training_frame=train, validation_frame=test)\n",
    "\n",
    "# print AUC\n",
    "print('GBM Test AUC = %.2f' % model.auc(valid=True))\n",
    "\n",
    "# uncomment to see model details\n",
    "# print(model) "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Display variable importance\n",
    "During training, the h2o GBM aggregates the improvement in error caused by each split in each decision tree across all the decision trees in the ensemble classifier. These values are attributed to the input variable used in each split and give an indication of the contribution each input variable makes toward the model's predictions. The variable importance ranking should be parsimonious with human domain knowledge and reasonable expectations. In this case, a customer's most recent payment behavior, `PAY_0`, is by far the most important variable followed by their second most recent payment, `PAY_2`, and third most recent payment, `PAY_3`, behavior. This result is well-aligned with business practices in credit lending: people who miss their most recent payments are likely to default soon."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f108ffa9cf8>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "model.varimp_plot()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 3. Train a decision tree surrogate model to describe GBM"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "A surrogate model is a simple model that is used to explain a complex model. One of the original references for surrogate models is available here: https://papers.nips.cc/paper/1152-extracting-tree-structured-representations-of-trained-networks.pdf. In this example, a single decision tree will be trained on the original inputs and predictions of the h2o GBM model and the tree will be visualized using special functionality in h2o and GraphViz. The variable importance, interactions, and decision paths displayed in the directed graph of the trained decision tree surrogate model are then assumed to be indicative of the internal mechanisms of the more complex GBM model, creating an approximate, overall flowchart for the GBM. There are few mathematical guarantees that the simple surrogate model is highly representative of the more complex GBM, but a recent preprint article has put forward ideas on strenghthening the theoretical relationship between surrogate models and more complex models: https://arxiv.org/pdf/1705.08504.pdf. Since surrogate models alone do not gaurantee accurate transparency, they will be used along with GBM variable importance and LOCO to build a cohesive narrative about the mechansims within the GBM. **Because many currently-available explanatory techniques are approximate, it is recommended that users employ several different explanatory techniques and trust only consisent results across techniques. Also, as of h2o 3.24, Shapley values are supported for h2o GBM. Use them instead of LOCO for any high-stakes application.**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Create dataset for surrogate model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To train a surrogate model, the predictions and original inputs of the complex model to be explained need to be in the same dataset. The test data is used here to see how the model behaves on holdout data, which should be closer to its behavior on new data than analyzing the surrogate model for the training inputs and predictions."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "gbm prediction progress: |████████████████████████████████████████████████| 100%\n"
     ]
    }
   ],
   "source": [
    "# cbind predictions to training frame\n",
    "# give them a nice name\n",
    "yhat = 'p_DEFAULT_NEXT_MONTH'\n",
    "preds1 = test['ID'].cbind(model.predict(test).drop(['predict', 'p0']))\n",
    "preds1.columns = ['ID', yhat]\n",
    "test_yhat = test.cbind(preds1[yhat])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Train single h2o decision tree\n",
    "A single decision tree is trained on the test  inputs and predictions. To simulate a single decision tree in h2o, the `H2ORandomForestEstimator` class is used, but only one tree is trained instead of a forest of decision trees. Setting the `mtry` parameter to `-2` tells the `H2ORandomForestEstimator` to consider all variables in all splits of a tree, instead of considering a random subset of columns. It is also recommended to set a random seed for reproducibility and to set `max_depth` to a lower number, say less than 6, so that the surrogate model will not become overly complex and hard to explain and understand. Once the tree is trained, a model optimized java object (MOJO) representation of the tree is saved. H2o provides a way to visualize the trained tree in detail using the MOJO and Graphviz."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "drf Model Build progress: |███████████████████████████████████████████████| 100%\n",
      "Generated MOJO path:\n",
      " /home/patrickh/workspace/interpretable_machine_learning_with_python/dt_surrogate_mojo.zip\n"
     ]
    }
   ],
   "source": [
    "model_id = 'dt_surrogate_mojo' # gives MOJO artifact a recognizable name\n",
    "\n",
    "# initialize single tree surrogate model\n",
    "surrogate = H2ORandomForestEstimator(ntrees=1,          # use only one tree\n",
    "                                     sample_rate=1,     # use all rows in that tree\n",
    "                                     mtries=-2,         # use all columns in that tree\n",
    "                                     max_depth=3,       # shallow trees are easier to understand\n",
    "                                     seed=12345,        # random seed for reproducibility\n",
    "                                     model_id=model_id) # gives MOJO artifact a recognizable name\n",
    "\n",
    "# train single tree surrogate model\n",
    "surrogate.train(x=X, y=yhat, training_frame=test_yhat)\n",
    "\n",
    "# persist MOJO (compiled, representation of trained model)\n",
    "# from which to generate plot of surrogate\n",
    "mojo_path = surrogate.download_mojo(path='.')\n",
    "print('Generated MOJO path:\\n', mojo_path)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Create GraphViz dot file\n",
    "GraphViz is an open source graph visualization tool. It is freely available from this url: http://www.graphviz.org/. To plot the trained decision tree surrogate model, a special h2o class, `PrintMojo`, is executed against the MOJO to create a GraphViz dot file representation of the tree."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Discovered H2O jar path:\n",
      " /home/patrickh/workspace/interpretable_machine_learning_with_python/env_iml/lib/python3.6/site-packages/h2o/backend/bin/h2o.jar\n",
      "\n",
      "Calling external process ...\n",
      "java -cp /home/patrickh/workspace/interpretable_machine_learning_with_python/env_iml/lib/python3.6/site-packages/h2o/backend/bin/h2o.jar hex.genmodel.tools.PrintMojo --tree 0 -i /home/patrickh/workspace/interpretable_machine_learning_with_python/dt_surrogate_mojo.zip -o dt_surrogate_mojo.gv --title Credit Card Default Decision Tree Surrogate\n"
     ]
    }
   ],
   "source": [
    "# title for plot\n",
    "title = 'Credit Card Default Decision Tree Surrogate'  \n",
    "\n",
    "# locate h2o jar\n",
    "hs = H2OLocalServer()\n",
    "h2o_jar_path = hs._find_jar()\n",
    "print('Discovered H2O jar path:\\n', h2o_jar_path)\n",
    "\n",
    "# construct command line call to generate graphviz version of \n",
    "# surrogate tree see for more information: \n",
    "# http://docs.h2o.ai/h2o/latest-stable/h2o-genmodel/javadoc/index.html\n",
    "gv_file_name = model_id + '.gv'\n",
    "gv_args = str('-cp ' + h2o_jar_path +\n",
    "              ' hex.genmodel.tools.PrintMojo --tree 0 -i '\n",
    "              + mojo_path + ' -o').split()\n",
    "gv_args.insert(0, 'java')\n",
    "gv_args.append(gv_file_name)\n",
    "if title is not None:\n",
    "    gv_args = gv_args + ['--title', title]\n",
    "    \n",
    "# call \n",
    "print()\n",
    "print('Calling external process ...')\n",
    "print(' '.join(gv_args))\n",
    "# if the line below is failing for you, try instead:\n",
    "# _ = subprocess.call(gv_args, shell=True)  \n",
    "_ = subprocess.call(gv_args)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Create PNG from GraphViz dot file and display\n",
    "Then a GraphViz command line tool is used to create a static PNG image from the dot file ... "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Calling external process ...\n",
      "dot -Tpng dt_surrogate_mojo.gv -o dt_surrogate_mojo.png\n"
     ]
    }
   ],
   "source": [
    "# construct call to generate PNG from \n",
    "# graphviz representation of the tree\n",
    "png_file_name = model_id + '.png'\n",
    "png_args = str('dot -Tpng ' + gv_file_name + ' -o ' + png_file_name)\n",
    "png_args = png_args.split()\n",
    "\n",
    "# call\n",
    "print('Calling external process ...')\n",
    "print(' '.join(png_args))\n",
    "# if the line below is failing for you, try instead:\n",
    "# _ = subprocess.call(png_args, shell=True)  \n",
    "_ = subprocess.call(png_args)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Display surrogate decision tree in notebook\n",
    "... and the image is displayed in the notebook."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# display in-notebook\n",
    "display(Image((png_file_name)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 4. Analyze surrogate model and compare to global GBM variable importance\n",
    "\n",
    "The displayed tree is comparable with the global GBM variable importance. A simple heuristic rule for variable importance in a decision tree relates to the depth and frequency at which a variable is split on in a tree: variables used higher in the tree and more frequently in the tree are more important. Most of the variables pictured in this tree also appear as highly important in the GBM variable importance plot. In both cases, `PAY_0` is appearing as crucially important, with other payment behavior variables following close behind. The surrogate decision tree enables users to understand and confirm not only what input variables are important, but also how their values contribute to model decisions. For instance, to fall into the lowest probability of default leaf node in the surrogate decision tree a customer must make their first and second payments in a timely fashion and then pay more than 1515.5 New Tiawanese Dollars for their fifth payment. Conversely, customers who miss their first, fifth, and third payments fall into the highest probability of default leaf node of the surrogate decision tree. It is also imperative to compare these results to domain knowledge and reasonable expectations. In this case, the global explanatory methods applied thus far tell a consisent and reasonable story about the GBM's behavior. If this was not so, steps should be taken to either reconcile or remove inconsistencies and unreasonable prediction behavior."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 5. Generate reason codes using the LOCO method \n",
    "\n",
    "Now that a solid understanding of global model behavior has been attained, local behavior for any given row of data and prediction can be analyzed and validated using LOCO. The LOCO method presented here is adapted from *Distribution-Free Predictive Inference for Regression* by Jing Lei et al., http://www.stat.cmu.edu/~ryantibs/papers/conformal.pdf. Here the local contribution of an input variable to a prediction for a single row of data is estimated by rescoring the GBM on that row one time for each input variable, each time leaving out one input variable (e.g., \"covariate\") by setting it to missing, and then subtracting the new score from the original score. By default, h2o scores missing data in decision trees by running them through the majority decision path. This means LOCO will be a numeric measure of how different the local contribution of an input variable is from the most common local contribution of that variable in the model. This variant of LOCO differs from the original method, in which one input variable is dropped from the model and the model is retrained without that variable. For nonlinear models, nonlinear dependencies can allow variables to nearly completely replace one another when a variable is dropped and the model is retrained. Hence, the approach of injecting missing values is used to estimate local contributions of input variables for nonlinear models here, as opposed to dropping a variable and retraining the model."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Calculate LOCO reason values for each row of the test set\n",
    "To implement LOCO, GBM model predicitions are calculated once for the test data and then again for each input variable, setting the entire input variable column to missing. Once the prediction without the variable is found for every row of data in the test set, that column vector of predictions on corrupted data can be subtracted from the column vector of predictions on the original, non-corrupted data to estimate the local contribution of that variable for each prediction in the test data. For better local accuracy and explainability, LOCO contributions are scaled such that contributions for each prediction plus the overall average of `DEFAULT_NEXT_MONTH` always sum to the model predictions."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Calculating LOCO contributions ...\n",
      "LOCO Progress: LIMIT_BAL (1/23) ...\n",
      "LOCO Progress: SEX (2/23) ...\n",
      "LOCO Progress: EDUCATION (3/23) ...\n",
      "LOCO Progress: MARRIAGE (4/23) ...\n",
      "LOCO Progress: AGE (5/23) ...\n",
      "LOCO Progress: PAY_0 (6/23) ...\n",
      "LOCO Progress: PAY_2 (7/23) ...\n",
      "LOCO Progress: PAY_3 (8/23) ...\n",
      "LOCO Progress: PAY_4 (9/23) ...\n",
      "LOCO Progress: PAY_5 (10/23) ...\n",
      "LOCO Progress: PAY_6 (11/23) ...\n",
      "LOCO Progress: BILL_AMT1 (12/23) ...\n",
      "LOCO Progress: BILL_AMT2 (13/23) ...\n",
      "LOCO Progress: BILL_AMT3 (14/23) ...\n",
      "LOCO Progress: BILL_AMT4 (15/23) ...\n",
      "LOCO Progress: BILL_AMT5 (16/23) ...\n",
      "LOCO Progress: BILL_AMT6 (17/23) ...\n",
      "LOCO Progress: PAY_AMT1 (18/23) ...\n",
      "LOCO Progress: PAY_AMT2 (19/23) ...\n",
      "LOCO Progress: PAY_AMT3 (20/23) ...\n",
      "LOCO Progress: PAY_AMT4 (21/23) ...\n",
      "LOCO Progress: PAY_AMT5 (22/23) ...\n",
      "LOCO Progress: PAY_AMT6 (23/23) ...\n",
      "\n",
      "Scaling contributions ...\n",
      "Done.\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>ID</th>\n",
       "      <th>p_DEFAULT_NEXT_MONTH</th>\n",
       "      <th>LIMIT_BAL</th>\n",
       "      <th>SEX</th>\n",
       "      <th>EDUCATION</th>\n",
       "      <th>MARRIAGE</th>\n",
       "      <th>AGE</th>\n",
       "      <th>PAY_0</th>\n",
       "      <th>PAY_2</th>\n",
       "      <th>PAY_3</th>\n",
       "      <th>...</th>\n",
       "      <th>BILL_AMT3</th>\n",
       "      <th>BILL_AMT4</th>\n",
       "      <th>BILL_AMT5</th>\n",
       "      <th>BILL_AMT6</th>\n",
       "      <th>PAY_AMT1</th>\n",
       "      <th>PAY_AMT2</th>\n",
       "      <th>PAY_AMT3</th>\n",
       "      <th>PAY_AMT4</th>\n",
       "      <th>PAY_AMT5</th>\n",
       "      <th>PAY_AMT6</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>4</td>\n",
       "      <td>0.144991</td>\n",
       "      <td>-0.079758</td>\n",
       "      <td>-0.000000</td>\n",
       "      <td>-0.00000</td>\n",
       "      <td>-0.009143</td>\n",
       "      <td>-0.001728</td>\n",
       "      <td>-0.000000</td>\n",
       "      <td>-0.000000</td>\n",
       "      <td>-0.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.002113</td>\n",
       "      <td>0.005768</td>\n",
       "      <td>-0.000000</td>\n",
       "      <td>-0.000000</td>\n",
       "      <td>-0.000000</td>\n",
       "      <td>-0.000000</td>\n",
       "      <td>-0.000000</td>\n",
       "      <td>-0.000000</td>\n",
       "      <td>-0.000000</td>\n",
       "      <td>-0.005340</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>8</td>\n",
       "      <td>0.128193</td>\n",
       "      <td>-0.020007</td>\n",
       "      <td>-0.000000</td>\n",
       "      <td>-0.00000</td>\n",
       "      <td>-0.000000</td>\n",
       "      <td>-0.000000</td>\n",
       "      <td>0.011403</td>\n",
       "      <td>-0.000000</td>\n",
       "      <td>0.045036</td>\n",
       "      <td>...</td>\n",
       "      <td>0.010062</td>\n",
       "      <td>-0.000000</td>\n",
       "      <td>-0.059467</td>\n",
       "      <td>0.015406</td>\n",
       "      <td>-0.028304</td>\n",
       "      <td>-0.036129</td>\n",
       "      <td>-0.067713</td>\n",
       "      <td>-0.000000</td>\n",
       "      <td>0.057314</td>\n",
       "      <td>-0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>10</td>\n",
       "      <td>0.179911</td>\n",
       "      <td>-0.024094</td>\n",
       "      <td>-0.002945</td>\n",
       "      <td>-0.00000</td>\n",
       "      <td>-0.000000</td>\n",
       "      <td>-0.009850</td>\n",
       "      <td>0.003778</td>\n",
       "      <td>-0.000000</td>\n",
       "      <td>0.005310</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.000000</td>\n",
       "      <td>-0.000000</td>\n",
       "      <td>-0.000000</td>\n",
       "      <td>-0.000000</td>\n",
       "      <td>-0.005015</td>\n",
       "      <td>-0.003077</td>\n",
       "      <td>-0.009657</td>\n",
       "      <td>0.019491</td>\n",
       "      <td>-0.000000</td>\n",
       "      <td>-0.004973</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>16</td>\n",
       "      <td>0.325205</td>\n",
       "      <td>0.012617</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.004043</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.027090</td>\n",
       "      <td>0.055267</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-0.003265</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.005403</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>17</td>\n",
       "      <td>0.408821</td>\n",
       "      <td>0.033188</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.00186</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.051523</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.001966</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.001247</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.008397</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 25 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   ID  p_DEFAULT_NEXT_MONTH  LIMIT_BAL       SEX  EDUCATION  MARRIAGE  \\\n",
       "0   4              0.144991  -0.079758 -0.000000   -0.00000 -0.009143   \n",
       "1   8              0.128193  -0.020007 -0.000000   -0.00000 -0.000000   \n",
       "2  10              0.179911  -0.024094 -0.002945   -0.00000 -0.000000   \n",
       "3  16              0.325205   0.012617  0.000000    0.00000  0.004043   \n",
       "4  17              0.408821   0.033188  0.000000    0.00186  0.000000   \n",
       "\n",
       "        AGE     PAY_0     PAY_2     PAY_3    ...     BILL_AMT3  BILL_AMT4  \\\n",
       "0 -0.001728 -0.000000 -0.000000 -0.000000    ...     -0.002113   0.005768   \n",
       "1 -0.000000  0.011403 -0.000000  0.045036    ...      0.010062  -0.000000   \n",
       "2 -0.009850  0.003778 -0.000000  0.005310    ...     -0.000000  -0.000000   \n",
       "3  0.000000  0.027090  0.055267  0.000000    ...      0.000000  -0.003265   \n",
       "4  0.000000  0.000000  0.000000  0.051523    ...     -0.001966   0.000000   \n",
       "\n",
       "   BILL_AMT5  BILL_AMT6  PAY_AMT1  PAY_AMT2  PAY_AMT3  PAY_AMT4  PAY_AMT5  \\\n",
       "0  -0.000000  -0.000000 -0.000000 -0.000000 -0.000000 -0.000000 -0.000000   \n",
       "1  -0.059467   0.015406 -0.028304 -0.036129 -0.067713 -0.000000  0.057314   \n",
       "2  -0.000000  -0.000000 -0.005015 -0.003077 -0.009657  0.019491 -0.000000   \n",
       "3   0.000000   0.000000  0.005403  0.000000  0.000000  0.000000  0.000000   \n",
       "4   0.000000   0.000000  0.000000  0.000000  0.000000  0.001247  0.000000   \n",
       "\n",
       "   PAY_AMT6  \n",
       "0 -0.005340  \n",
       "1 -0.000000  \n",
       "2 -0.004973  \n",
       "3  0.000000  \n",
       "4  0.008397  \n",
       "\n",
       "[5 rows x 25 columns]"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "h2o.no_progress() # turn off h2o gratuitous progress bars\n",
    "\n",
    "# create set of original predictions and row ID\n",
    "preds2 = test['ID'].cbind(model.predict(test).drop(['predict', 'p0']))\n",
    "preds2.columns = ['ID', yhat]\n",
    "\n",
    "# calculate LOCO for each variable\n",
    "print('Calculating LOCO contributions ...')\n",
    "for k, i in enumerate(X):\n",
    "\n",
    "    # train and predict with x_i set to missing\n",
    "    test_loco = h2o.deep_copy(test, 'test_loco')\n",
    "    test_loco[i] = np.nan\n",
    "    preds_loco = model.predict(test_loco).drop(['predict','p0'])\n",
    "    \n",
    "    # create a new, named column for the LOCO prediction\n",
    "    preds_loco.columns = [i]\n",
    "    preds2 = preds2.cbind(preds_loco)\n",
    "    \n",
    "    # subtract the LOCO prediction from the original prediction\n",
    "    preds2[i] = preds2[yhat] - preds2[i]\n",
    "    \n",
    "    # update progress\n",
    "    print('LOCO Progress: ' + i + ' (' + str(k+1) + '/' + str(len(X)) + ') ...')\n",
    "    \n",
    "# scale contributions to sum to yhat - y_0\n",
    "print('\\nScaling contributions ...')\n",
    "\n",
    "y_0 = test[y].mean()[0]\n",
    "preds2_pd = preds2.as_data_frame()\n",
    "pred_ = preds2_pd[yhat]\n",
    "scaler = (pred_ - y_0) / preds2_pd[X].sum(axis=1)\n",
    "preds2_pd[X] = preds2_pd[X].multiply(scaler, axis=0)   \n",
    "\n",
    "print('Done.')  \n",
    "\n",
    "preds2_pd.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The numeric LOCO values in each column are an estimate of how much each variable contributed to each prediction. LOCO can indicate how a variable and its values were weighted in any given decision by the model. These values are crucially important for machine learning interpretability and are related to \"local feature importance\", \"reason codes\", or \"turn-down codes.\" The latter phrases are borrowed from credit scoring. Credit lenders in the U.S. must provide reasons for automatically rejecting a credit application. Reason codes can be easily extracted from LOCO local variable contribution values by simply ranking the variables that played the largest role in any given decision."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Helper function for finding percentile indices\n",
    "The function below finds and returns the row indices for the minimum, the maximum, and the deciles of one column in terms of another, in this case the model predictions (`p_DEFAULT_NEXT_MONTH`) and the row identifier (`ID`), respectively. These indices are used as a starting point for finding potentially interesting predictions. Outlying predictions found through residual analysis is another group of potentially interesting local predictions to analyze with LOCO."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{0: 28716,\n",
       " 99: 29116,\n",
       " 10: 8942,\n",
       " 20: 28257,\n",
       " 30: 4074,\n",
       " 40: 13411,\n",
       " 50: 16633,\n",
       " 60: 2402,\n",
       " 70: 19769,\n",
       " 80: 25069,\n",
       " 90: 21372}"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def get_percentile_dict(yhat, id_, frame):\n",
    "\n",
    "    \"\"\" Returns the minimum, maximum, and percentiles of a column, yhat, \n",
    "        as the indices based on another column id_.\n",
    "    \n",
    "    Args:\n",
    "        yhat: Column in which to find percentiles.\n",
    "        id_: Id column that stores indices for percentiles of yhat.\n",
    "        frame: H2OFrame containing yhat and id_. \n",
    "    \n",
    "    Returns:\n",
    "        Dictionary of percentile values and index column values.\n",
    "    \n",
    "    \"\"\"\n",
    "    \n",
    "    # convert to Pandas and sort \n",
    "    sort_df = preds2_pd.copy(deep=True)\n",
    "    sort_df.sort_values(yhat, inplace=True)\n",
    "    sort_df.reset_index(inplace=True)\n",
    "\n",
    "    # find top and bottom percentiles\n",
    "    percentiles_dict = {}\n",
    "    percentiles_dict[0] = sort_df.loc[0, id_]\n",
    "    percentiles_dict[99] = sort_df.loc[sort_df.shape[0]-1, id_]\n",
    "    inc = sort_df.shape[0]//10\n",
    "    \n",
    "    # find 10th-90th percentiles    \n",
    "    for i in range(1, 10):\n",
    "        percentiles_dict[i * 10] = sort_df.loc[i * inc,  id_]\n",
    "\n",
    "    return percentiles_dict\n",
    "\n",
    "# display percentiles dictionary\n",
    "# ID values for rows\n",
    "# from lowest prediction \n",
    "# to highest prediction\n",
    "percentile_dict = get_percentile_dict(yhat, 'ID', preds2_pd)\n",
    "percentile_dict"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Plot some reason codes for a risky customer\n",
    "Investigating customers with very high or low predicted probabilities to determine if their local explanations justify their extreme predictions is typically a productive exercise in boundary testing, model debugging, and validation. Reason codes are generated for the customer with the highest probability of default in the test data set below in cell 18, but LOCO can create local explanations for any or all rows in the training or test datasets, and on new data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f108e8d7b70>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# select single customer\n",
    "# convert to Pandas\n",
    "# drop prediction and row ID\n",
    "\n",
    "risky_loco = preds2_pd[preds2_pd['ID'] == int(percentile_dict[99])].drop(['ID', yhat], axis=1)   \n",
    "\n",
    "# transpose into column vector and sort        \n",
    "risky_loco = risky_loco.T.sort_values(by=8674, ascending=False)[:5]\n",
    "\n",
    "# plot\n",
    "_ = risky_loco.plot(kind='bar', \n",
    "                    title='Top Five Reason Codes for a Risky Customer\\n', \n",
    "                    legend=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "For the customer in the test dataset that the GBM predicts as most likely to default, the most important input variables in the prediction are, in descending order, `PAY_0`, `PAY_6`, `PAY_3`, `PAY_5`, and `AGE`."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Display customer in question "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The local contributions for this customer appear reasonable, especially when considering her payment information. Her most recent payment was 3 months late and her payment for 6 months previous was 4 months late, so it's logical that these would weigh heavily into the model's prediction for default for this customer."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table>\n",
       "<thead>\n",
       "<tr><th style=\"text-align: right;\">   ID</th><th style=\"text-align: right;\">  LIMIT_BAL</th><th>SEX   </th><th>EDUCATION  </th><th>MARRIAGE  </th><th style=\"text-align: right;\">  AGE</th><th>PAY_0        </th><th>PAY_2        </th><th>PAY_3        </th><th>PAY_4        </th><th>PAY_5        </th><th>PAY_6        </th><th style=\"text-align: right;\">  BILL_AMT1</th><th style=\"text-align: right;\">  BILL_AMT2</th><th style=\"text-align: right;\">  BILL_AMT3</th><th style=\"text-align: right;\">  BILL_AMT4</th><th style=\"text-align: right;\">  BILL_AMT5</th><th style=\"text-align: right;\">  BILL_AMT6</th><th style=\"text-align: right;\">  PAY_AMT1</th><th style=\"text-align: right;\">  PAY_AMT2</th><th style=\"text-align: right;\">  PAY_AMT3</th><th style=\"text-align: right;\">  PAY_AMT4</th><th style=\"text-align: right;\">  PAY_AMT5</th><th style=\"text-align: right;\">  PAY_AMT6</th><th style=\"text-align: right;\">  DEFAULT_NEXT_MONTH</th><th style=\"text-align: right;\">  p_DEFAULT_NEXT_MONTH</th></tr>\n",
       "</thead>\n",
       "<tbody>\n",
       "<tr><td style=\"text-align: right;\">29116</td><td style=\"text-align: right;\">      20000</td><td>female</td><td>university </td><td>married   </td><td style=\"text-align: right;\">   59</td><td>3 month delay</td><td>2 month delay</td><td>3 month delay</td><td>2 month delay</td><td>2 month delay</td><td>4 month delay</td><td style=\"text-align: right;\">       8803</td><td style=\"text-align: right;\">      11137</td><td style=\"text-align: right;\">      10672</td><td style=\"text-align: right;\">      11201</td><td style=\"text-align: right;\">      12721</td><td style=\"text-align: right;\">      11946</td><td style=\"text-align: right;\">      2800</td><td style=\"text-align: right;\">         0</td><td style=\"text-align: right;\">      1000</td><td style=\"text-align: right;\">      2000</td><td style=\"text-align: right;\">         0</td><td style=\"text-align: right;\">         0</td><td style=\"text-align: right;\">                   1</td><td style=\"text-align: right;\">              0.895285</td></tr>\n",
       "</tbody>\n",
       "</table>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": []
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_yhat[test_yhat['ID'] == int(percentile_dict[99]), :] # helps understand reason codes"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To generate reason codes for the model's decision, the locally important variable and its value are used together. If this customer was denied future credit based on this model and data, the top five LOCO-based reason codes for the automated decision would be:\n",
    "\n",
    "1. Most recent payment is 3 months delayed.\n",
    "2. 6th most recent payment is 4 months delayed.\n",
    "3. 3rd most recent payment is 3 months delayed.\n",
    "4. 5th most recent payment is 2 months delayed.\n",
    "5. Customer age is 59. \n",
    "\n",
    "(Of course, in many places, variables like `AGE` and `SEX` cannot, and should not, be used in credit lending or other high-stakes decisions. For a slightly more careful treatment of GBM in a fair lending context, see: https://github.com/jphall663/interpretable_machine_learning_with_python/blob/master/dia.ipynb)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 6. Bonus: Generate ensemble LOCO reason codes for greater explanation stability\n",
    "Just like predictions from high variance, nonlinear models, *explanations* derived from machine learning models can be unstable. One general way to decrease variance is to ensemble the results of many models. The last section of this notebook puts forward a simple approach to creating ensemble explanations."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Train multiple models\n",
    "To create ensemble explanations, several accurate models are trained. The models and their predictions on the test data are stored in Python lists."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training Progress: model 1/10, AUC = 0.7813 ...\n",
      "Training Progress: model 2/10, AUC = 0.7803 ...\n",
      "Training Progress: model 3/10, AUC = 0.7787 ...\n",
      "Training Progress: model 4/10, AUC = 0.7826 ...\n",
      "Training Progress: model 5/10, AUC = 0.7804 ...\n",
      "Training Progress: model 6/10, AUC = 0.7800 ...\n",
      "Training Progress: model 7/10, AUC = 0.7802 ...\n",
      "Training Progress: model 8/10, AUC = 0.7799 ...\n",
      "Training Progress: model 9/10, AUC = 0.7796 ...\n",
      "Training Progress: model 10/10, AUC = 0.7811 ...\n",
      "Done.\n"
     ]
    }
   ],
   "source": [
    "n_models = 10 # select number of models\n",
    "\n",
    "# lists for holding models and predictions\n",
    "models = []\n",
    "pred_frames = []\n",
    "\n",
    "for i in range(0, n_models):\n",
    "\n",
    "    # initialize and store models\n",
    "    models.append(H2OGradientBoostingEstimator(ntrees=150,\n",
    "                                               max_depth=4,\n",
    "                                               sample_rate=0.9 - ((i + 1)*0.01),     # perturb sample rate\n",
    "                                               col_sample_rate=0.9 - ((i + 1)*0.01), # perturb column sample rate\n",
    "                                               #balance_classes=True,                # sample to balance 0/1 distribution of target - helps LOCO\n",
    "                                               stopping_rounds=5,                    # stop if validation error does not decrease for 5 iterations (trees)\n",
    "                                               seed=i + 1))                          # new random seed for each model\n",
    "    \n",
    "    # train models\n",
    "    models[i].train(y=y, x=X, training_frame=train, validation_frame=test)\n",
    "    \n",
    "    # store predictions\n",
    "    pred_frames.append(test['ID'].cbind(models[i].predict(test).drop(['predict','p0'])))\n",
    "    pred_frames[i].columns = ['ID', yhat]\n",
    "    \n",
    "    # update progress\n",
    "    print('Training Progress: model %d/%d, AUC = %.4f ...' % (i + 1, n_models, models[i].auc(valid=True)))\n",
    "\n",
    "print('Done.')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Calculate LOCO for each model\n",
    "LOCO is calculated on the test data for each model, each input, and each row of data in the test set using the stored models and predictions."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "LOCO Progress: model 1/10 ...\n",
      "LOCO Progress: model 2/10 ...\n",
      "LOCO Progress: model 3/10 ...\n",
      "LOCO Progress: model 4/10 ...\n",
      "LOCO Progress: model 5/10 ...\n",
      "LOCO Progress: model 6/10 ...\n",
      "LOCO Progress: model 7/10 ...\n",
      "LOCO Progress: model 8/10 ...\n",
      "LOCO Progress: model 9/10 ...\n",
      "LOCO Progress: model 10/10 ...\n",
      "Done.\n"
     ]
    }
   ],
   "source": [
    "# for each new model ...\n",
    "for k, model in enumerate(models):\n",
    "\n",
    "    # calculate LOCO for each input variable \n",
    "    for i in X:\n",
    "\n",
    "        # train and predict with Xi set to missing\n",
    "        test_loco = h2o.deep_copy(test, 'test_loco')\n",
    "        test_loco[i] = np.nan\n",
    "        preds_loco = model.predict(test_loco).drop(['predict','p0'])\n",
    "\n",
    "        # create a new, named column for the LOCO prediction\n",
    "        preds_loco.columns = [i]\n",
    "        pred_frames[k] = pred_frames[k].cbind(preds_loco)\n",
    "\n",
    "        # subtract the LOCO prediction from the original prediction\n",
    "        pred_frames[k][i] = pred_frames[k][yhat] - pred_frames[k][i]\n",
    "        \n",
    "    # update progress    \n",
    "    print('LOCO Progress: model %d/%d ...' % (k + 1, n_models))\n",
    "\n",
    "print('Done.')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Collect LOCO values for each model for a risky customer\n",
    "To create ensemble explanations for a single row, the LOCO values for each variable in the row are averaged across all models. Single-model and mean LOCO values for the most risky person in the test set are displayed below. Notice that even slight changes in model specifications can result in different explanations. For example, the local contribution of `PAY_0` for the riskiest customer ranges from 0.13 to 0.23 across the 10 models in the table below. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Loco 1</th>\n",
       "      <th>Loco 2</th>\n",
       "      <th>Loco 3</th>\n",
       "      <th>Loco 4</th>\n",
       "      <th>Loco 5</th>\n",
       "      <th>Loco 6</th>\n",
       "      <th>Loco 7</th>\n",
       "      <th>Loco 8</th>\n",
       "      <th>Loco 9</th>\n",
       "      <th>Loco 10</th>\n",
       "      <th>Mean Local Importance</th>\n",
       "      <th>Scaled Mean Local Importance</th>\n",
       "      <th>Std. Dev. Local Importance</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>LIMIT_BAL</th>\n",
       "      <td>0.013205</td>\n",
       "      <td>0.011040</td>\n",
       "      <td>0.012483</td>\n",
       "      <td>0.001345</td>\n",
       "      <td>0.015150</td>\n",
       "      <td>-0.006688</td>\n",
       "      <td>-0.002158</td>\n",
       "      <td>-0.009428</td>\n",
       "      <td>-0.002005</td>\n",
       "      <td>-0.005503</td>\n",
       "      <td>0.002744</td>\n",
       "      <td>0.004116</td>\n",
       "      <td>0.008836</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SEX</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>EDUCATION</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MARRIAGE</th>\n",
       "      <td>0.051186</td>\n",
       "      <td>0.023304</td>\n",
       "      <td>0.049574</td>\n",
       "      <td>0.044072</td>\n",
       "      <td>0.034987</td>\n",
       "      <td>0.083513</td>\n",
       "      <td>0.053444</td>\n",
       "      <td>0.034809</td>\n",
       "      <td>0.051611</td>\n",
       "      <td>0.097880</td>\n",
       "      <td>0.052438</td>\n",
       "      <td>0.078652</td>\n",
       "      <td>0.021392</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>AGE</th>\n",
       "      <td>0.022646</td>\n",
       "      <td>0.084028</td>\n",
       "      <td>0.015365</td>\n",
       "      <td>0.019450</td>\n",
       "      <td>0.047277</td>\n",
       "      <td>0.008411</td>\n",
       "      <td>0.019311</td>\n",
       "      <td>0.019194</td>\n",
       "      <td>0.064969</td>\n",
       "      <td>0.047981</td>\n",
       "      <td>0.034863</td>\n",
       "      <td>0.052292</td>\n",
       "      <td>0.023673</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>PAY_0</th>\n",
       "      <td>0.199285</td>\n",
       "      <td>0.136971</td>\n",
       "      <td>0.155368</td>\n",
       "      <td>0.123913</td>\n",
       "      <td>0.103977</td>\n",
       "      <td>0.094563</td>\n",
       "      <td>0.129808</td>\n",
       "      <td>0.166523</td>\n",
       "      <td>0.163303</td>\n",
       "      <td>0.118950</td>\n",
       "      <td>0.139266</td>\n",
       "      <td>0.208887</td>\n",
       "      <td>0.030281</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>PAY_2</th>\n",
       "      <td>0.002721</td>\n",
       "      <td>0.023228</td>\n",
       "      <td>0.029913</td>\n",
       "      <td>0.064136</td>\n",
       "      <td>0.047839</td>\n",
       "      <td>0.003387</td>\n",
       "      <td>0.044296</td>\n",
       "      <td>0.002752</td>\n",
       "      <td>0.021887</td>\n",
       "      <td>0.028216</td>\n",
       "      <td>0.026837</td>\n",
       "      <td>0.040254</td>\n",
       "      <td>0.019742</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>PAY_3</th>\n",
       "      <td>0.068494</td>\n",
       "      <td>0.093103</td>\n",
       "      <td>0.047630</td>\n",
       "      <td>0.046408</td>\n",
       "      <td>0.044800</td>\n",
       "      <td>0.016914</td>\n",
       "      <td>0.038656</td>\n",
       "      <td>0.046357</td>\n",
       "      <td>0.088166</td>\n",
       "      <td>0.054055</td>\n",
       "      <td>0.054458</td>\n",
       "      <td>0.081683</td>\n",
       "      <td>0.021809</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>PAY_4</th>\n",
       "      <td>0.030388</td>\n",
       "      <td>0.053298</td>\n",
       "      <td>0.039892</td>\n",
       "      <td>-0.009098</td>\n",
       "      <td>0.021445</td>\n",
       "      <td>0.022628</td>\n",
       "      <td>0.043056</td>\n",
       "      <td>0.038197</td>\n",
       "      <td>0.063759</td>\n",
       "      <td>0.021841</td>\n",
       "      <td>0.032541</td>\n",
       "      <td>0.048808</td>\n",
       "      <td>0.019174</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>PAY_5</th>\n",
       "      <td>0.064508</td>\n",
       "      <td>0.051064</td>\n",
       "      <td>0.033604</td>\n",
       "      <td>0.069855</td>\n",
       "      <td>0.024448</td>\n",
       "      <td>0.038036</td>\n",
       "      <td>0.031212</td>\n",
       "      <td>0.020676</td>\n",
       "      <td>0.063064</td>\n",
       "      <td>0.072998</td>\n",
       "      <td>0.046947</td>\n",
       "      <td>0.070416</td>\n",
       "      <td>0.018681</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>PAY_6</th>\n",
       "      <td>0.030094</td>\n",
       "      <td>0.033554</td>\n",
       "      <td>0.027220</td>\n",
       "      <td>-0.023089</td>\n",
       "      <td>0.005430</td>\n",
       "      <td>0.005204</td>\n",
       "      <td>0.022503</td>\n",
       "      <td>0.029180</td>\n",
       "      <td>0.028095</td>\n",
       "      <td>0.030959</td>\n",
       "      <td>0.018915</td>\n",
       "      <td>0.028371</td>\n",
       "      <td>0.017000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BILL_AMT1</th>\n",
       "      <td>0.000933</td>\n",
       "      <td>0.036190</td>\n",
       "      <td>0.002133</td>\n",
       "      <td>0.025375</td>\n",
       "      <td>0.001071</td>\n",
       "      <td>0.004628</td>\n",
       "      <td>0.018029</td>\n",
       "      <td>0.004658</td>\n",
       "      <td>0.021524</td>\n",
       "      <td>0.027100</td>\n",
       "      <td>0.014164</td>\n",
       "      <td>0.021245</td>\n",
       "      <td>0.012328</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BILL_AMT2</th>\n",
       "      <td>0.008804</td>\n",
       "      <td>0.001631</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.031767</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.008615</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.005082</td>\n",
       "      <td>0.007622</td>\n",
       "      <td>0.009515</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BILL_AMT3</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>-0.001008</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-0.001648</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-0.000266</td>\n",
       "      <td>-0.000398</td>\n",
       "      <td>0.000550</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BILL_AMT4</th>\n",
       "      <td>-0.001370</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-0.010070</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-0.005586</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-0.001738</td>\n",
       "      <td>-0.001876</td>\n",
       "      <td>-0.002814</td>\n",
       "      <td>0.003198</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BILL_AMT5</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.003339</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.004020</td>\n",
       "      <td>-0.010965</td>\n",
       "      <td>0.009204</td>\n",
       "      <td>0.000560</td>\n",
       "      <td>0.000840</td>\n",
       "      <td>0.004787</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BILL_AMT6</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.006478</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000648</td>\n",
       "      <td>0.000972</td>\n",
       "      <td>0.001943</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>PAY_AMT1</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.004307</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.003919</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000823</td>\n",
       "      <td>0.001234</td>\n",
       "      <td>0.001648</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>PAY_AMT2</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.001541</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-0.001350</td>\n",
       "      <td>0.004503</td>\n",
       "      <td>0.002450</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.001922</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000985</td>\n",
       "      <td>0.001005</td>\n",
       "      <td>0.001508</td>\n",
       "      <td>0.001582</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>PAY_AMT3</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.005040</td>\n",
       "      <td>0.000504</td>\n",
       "      <td>0.000756</td>\n",
       "      <td>0.001512</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>PAY_AMT4</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-0.006528</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-0.000653</td>\n",
       "      <td>-0.000979</td>\n",
       "      <td>0.001959</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>PAY_AMT5</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.014605</td>\n",
       "      <td>0.016906</td>\n",
       "      <td>0.016336</td>\n",
       "      <td>0.001440</td>\n",
       "      <td>0.004975</td>\n",
       "      <td>0.029793</td>\n",
       "      <td>0.045810</td>\n",
       "      <td>0.019439</td>\n",
       "      <td>0.047491</td>\n",
       "      <td>0.019679</td>\n",
       "      <td>0.029517</td>\n",
       "      <td>0.015936</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>PAY_AMT6</th>\n",
       "      <td>-0.000925</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-0.009342</td>\n",
       "      <td>-0.000800</td>\n",
       "      <td>-0.000556</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-0.001162</td>\n",
       "      <td>-0.001743</td>\n",
       "      <td>0.002749</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             Loco 1    Loco 2    Loco 3    Loco 4    Loco 5    Loco 6  \\\n",
       "LIMIT_BAL  0.013205  0.011040  0.012483  0.001345  0.015150 -0.006688   \n",
       "SEX        0.000000  0.000000  0.000000  0.000000  0.000000  0.000000   \n",
       "EDUCATION  0.000000  0.000000  0.000000  0.000000  0.000000  0.000000   \n",
       "MARRIAGE   0.051186  0.023304  0.049574  0.044072  0.034987  0.083513   \n",
       "AGE        0.022646  0.084028  0.015365  0.019450  0.047277  0.008411   \n",
       "PAY_0      0.199285  0.136971  0.155368  0.123913  0.103977  0.094563   \n",
       "PAY_2      0.002721  0.023228  0.029913  0.064136  0.047839  0.003387   \n",
       "PAY_3      0.068494  0.093103  0.047630  0.046408  0.044800  0.016914   \n",
       "PAY_4      0.030388  0.053298  0.039892 -0.009098  0.021445  0.022628   \n",
       "PAY_5      0.064508  0.051064  0.033604  0.069855  0.024448  0.038036   \n",
       "PAY_6      0.030094  0.033554  0.027220 -0.023089  0.005430  0.005204   \n",
       "BILL_AMT1  0.000933  0.036190  0.002133  0.025375  0.001071  0.004628   \n",
       "BILL_AMT2  0.008804  0.001631  0.000000  0.000000  0.031767  0.000000   \n",
       "BILL_AMT3  0.000000 -0.001008  0.000000  0.000000  0.000000  0.000000   \n",
       "BILL_AMT4 -0.001370  0.000000  0.000000  0.000000  0.000000 -0.010070   \n",
       "BILL_AMT5  0.000000  0.000000  0.000000  0.003339  0.000000  0.000000   \n",
       "BILL_AMT6  0.000000  0.000000  0.000000  0.000000  0.006478  0.000000   \n",
       "PAY_AMT1   0.000000  0.004307  0.000000  0.000000  0.000000  0.003919   \n",
       "PAY_AMT2   0.000000  0.001541  0.000000 -0.001350  0.004503  0.002450   \n",
       "PAY_AMT3   0.000000  0.000000  0.000000  0.000000  0.000000  0.000000   \n",
       "PAY_AMT4   0.000000  0.000000  0.000000 -0.006528  0.000000  0.000000   \n",
       "PAY_AMT5   0.000000  0.014605  0.016906  0.016336  0.001440  0.004975   \n",
       "PAY_AMT6  -0.000925  0.000000  0.000000 -0.009342 -0.000800 -0.000556   \n",
       "\n",
       "             Loco 7    Loco 8    Loco 9   Loco 10  Mean Local Importance  \\\n",
       "LIMIT_BAL -0.002158 -0.009428 -0.002005 -0.005503               0.002744   \n",
       "SEX        0.000000  0.000000  0.000000  0.000000               0.000000   \n",
       "EDUCATION  0.000000  0.000000  0.000000  0.000000               0.000000   \n",
       "MARRIAGE   0.053444  0.034809  0.051611  0.097880               0.052438   \n",
       "AGE        0.019311  0.019194  0.064969  0.047981               0.034863   \n",
       "PAY_0      0.129808  0.166523  0.163303  0.118950               0.139266   \n",
       "PAY_2      0.044296  0.002752  0.021887  0.028216               0.026837   \n",
       "PAY_3      0.038656  0.046357  0.088166  0.054055               0.054458   \n",
       "PAY_4      0.043056  0.038197  0.063759  0.021841               0.032541   \n",
       "PAY_5      0.031212  0.020676  0.063064  0.072998               0.046947   \n",
       "PAY_6      0.022503  0.029180  0.028095  0.030959               0.018915   \n",
       "BILL_AMT1  0.018029  0.004658  0.021524  0.027100               0.014164   \n",
       "BILL_AMT2  0.008615  0.000000  0.000000  0.000000               0.005082   \n",
       "BILL_AMT3  0.000000  0.000000 -0.001648  0.000000              -0.000266   \n",
       "BILL_AMT4  0.000000 -0.005586  0.000000 -0.001738              -0.001876   \n",
       "BILL_AMT5  0.000000  0.004020 -0.010965  0.009204               0.000560   \n",
       "BILL_AMT6  0.000000  0.000000  0.000000  0.000000               0.000648   \n",
       "PAY_AMT1   0.000000  0.000000  0.000000  0.000000               0.000823   \n",
       "PAY_AMT2   0.000000  0.001922  0.000000  0.000985               0.001005   \n",
       "PAY_AMT3   0.000000  0.000000  0.000000  0.005040               0.000504   \n",
       "PAY_AMT4   0.000000  0.000000  0.000000  0.000000              -0.000653   \n",
       "PAY_AMT5   0.029793  0.045810  0.019439  0.047491               0.019679   \n",
       "PAY_AMT6   0.000000  0.000000  0.000000  0.000000              -0.001162   \n",
       "\n",
       "           Scaled Mean Local Importance  Std. Dev. Local Importance  \n",
       "LIMIT_BAL                      0.004116                    0.008836  \n",
       "SEX                            0.000000                    0.000000  \n",
       "EDUCATION                      0.000000                    0.000000  \n",
       "MARRIAGE                       0.078652                    0.021392  \n",
       "AGE                            0.052292                    0.023673  \n",
       "PAY_0                          0.208887                    0.030281  \n",
       "PAY_2                          0.040254                    0.019742  \n",
       "PAY_3                          0.081683                    0.021809  \n",
       "PAY_4                          0.048808                    0.019174  \n",
       "PAY_5                          0.070416                    0.018681  \n",
       "PAY_6                          0.028371                    0.017000  \n",
       "BILL_AMT1                      0.021245                    0.012328  \n",
       "BILL_AMT2                      0.007622                    0.009515  \n",
       "BILL_AMT3                     -0.000398                    0.000550  \n",
       "BILL_AMT4                     -0.002814                    0.003198  \n",
       "BILL_AMT5                      0.000840                    0.004787  \n",
       "BILL_AMT6                      0.000972                    0.001943  \n",
       "PAY_AMT1                       0.001234                    0.001648  \n",
       "PAY_AMT2                       0.001508                    0.001582  \n",
       "PAY_AMT3                       0.000756                    0.001512  \n",
       "PAY_AMT4                      -0.000979                    0.001959  \n",
       "PAY_AMT5                       0.029517                    0.015936  \n",
       "PAY_AMT6                      -0.001743                    0.002749  "
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# holds predictions for a specific row \n",
    "risky_loco_frames = []\n",
    "\n",
    "# column names for Pandas DataFrame of combined LOCO prediction\n",
    "col_names = ['Loco ' + str(i) for i in range(1, n_models + 1)]\n",
    "\n",
    "# for each new model ...\n",
    "for i in range(0, n_models):\n",
    "    \n",
    "    # collect LOCO for that model and a specific row \n",
    "    # as a column vector in a Pandas DataFrame\n",
    "    preds = pred_frames[i]\n",
    "    risky_loco_frames.append(preds[preds['ID'] == int(percentile_dict[99]), :] # row for risky person\n",
    "                             .as_data_frame()                                  # convert to Pandas\n",
    "                             .drop(['ID', yhat], axis=1)                       # drop predictions and row ID\n",
    "                             .T)                                               # Transpose into column vector\n",
    "\n",
    "# bind LOCO for each row as column vectors \n",
    "# into the same Pandas DataFrame\n",
    "loco_ensemble = pd.concat(risky_loco_frames, axis=1) \n",
    "\n",
    "# update column names\n",
    "loco_ensemble.columns = col_names\n",
    "\n",
    "# mean local importance across models\n",
    "loco_ensemble['Mean Local Importance'] = loco_ensemble.mean(axis=1)\n",
    "\n",
    "# scale contribs\n",
    "scaler = (test_yhat[test_yhat['ID'] == int(percentile_dict[99]), yhat] - y_0) /\\\n",
    "    (loco_ensemble['Mean Local Importance'].sum())\n",
    "loco_ensemble['Scaled Mean Local Importance'] = loco_ensemble['Mean Local Importance'] * scaler[0, 0]\n",
    "\n",
    "# std deviation\n",
    "loco_ensemble['Std. Dev. Local Importance'] = loco_ensemble\\\n",
    "                                              .drop('Scaled Mean Local Importance', axis=1)\\\n",
    "                                              .std(axis=1)\n",
    "        \n",
    "# display\n",
    "loco_ensemble"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Plot some mean reason codes for a risky customer\n",
    "Taking mean explanations across multiple models leads to reason codes somewhat different from the reason codes produced by a single model. Mean reason codes may be more stable, they represent explanations from several models, and they may take practicioners a step closer to using machine learning models to make inferential conclusions about phenomena represented in the training or test data, instead of simply providing an approximate explanation of a single model's decision processes. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f10836db400>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "risky_mean_loco = loco_ensemble['Mean Local Importance'].sort_values(ascending=False)[:5]\n",
    "_ = risky_mean_loco.plot(kind='bar', \n",
    "                         title='Top Five Reason Codes for a Risky Customer\\n', \n",
    "                         color='b',\n",
    "                         legend=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Shutdown H2O\n",
    "After using h2o, it's typically best to shut it down. However, before doing so, users should ensure that they have saved any h2o data structures, such as models and H2OFrames, or scoring artifacts, such as POJOs and MOJOs."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "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_ae25 closed.\n"
     ]
    }
   ],
   "source": [
    "# be careful, this can erase your work!\n",
    "h2o.cluster().shutdown(prompt=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Summary\n",
    "\n",
    "In this notebook, a complex GBM classifier was trained to predict credit card defaults and explained at a global scale with a decision tree surrogate model and explained at a local scale with LOCO. An ensemble LOCO approach was also introduced to stabilize approximate explanations. The decision tree surrogate creates an overall approximate flowchart for the GBM's decision processes and LOCO can be used to create reason codes for each model prediction. All of these techniques enhance the transparency of the complex model, which in turn enables greater accountability for the model's predictions. These techniques should generalize well for many types of business and research problems, enabling you to train a complex GBM model and explain it to your colleagues, bosses, and potentially, external regulators. "
   ]
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}