forage-anz-r.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Load libraries"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Registered S3 methods overwritten by 'ggplot2':\n",
      "  method         from \n",
      "  [.quosures     rlang\n",
      "  c.quosures     rlang\n",
      "  print.quosures rlang\n",
      "Registered S3 method overwritten by 'rvest':\n",
      "  method            from\n",
      "  read_xml.response xml2\n",
      "-- Attaching packages --------------------------------------- tidyverse 1.2.1 --\n",
      "v ggplot2 3.1.1       v purrr   0.3.2  \n",
      "v tibble  2.1.1       v dplyr   0.8.0.1\n",
      "v tidyr   0.8.3       v stringr 1.4.0  \n",
      "v readr   1.3.1       v forcats 0.4.0  \n",
      "-- Conflicts ------------------------------------------ tidyverse_conflicts() --\n",
      "x dplyr::filter() masks stats::filter()\n",
      "x dplyr::lag()    masks stats::lag()\n",
      "\n",
      "Attaching package: 'lubridate'\n",
      "\n",
      "The following object is masked from 'package:base':\n",
      "\n",
      "    date\n",
      "\n",
      "Warning message:\n",
      "\"package 'Metrics' was built under R version 3.6.3\"Warning message:\n",
      "\"package 'rpart' was built under R version 3.6.3\""
     ]
    }
   ],
   "source": [
    "library(tidyverse)\n",
    "library(readxl)\n",
    "library(lubridate)\n",
    "library(Metrics)\n",
    "library(rpart)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Load transaction dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Warning message in read_fun(path = enc2native(normalizePath(path)), sheet_i = sheet, :\n",
      "\"Expecting numeric in C3052 / R3052C3: got 'THE DISCOUNT CHEMIST GROUP'\"Warning message in read_fun(path = enc2native(normalizePath(path)), sheet_i = sheet, :\n",
      "\"Expecting numeric in C4360 / R4360C3: got 'LAND WATER & PLANNING East Melbourne'\""
     ]
    },
    {
     "data": {
      "text/html": [
       "<table>\n",
       "<thead><tr><th scope=col>status</th><th scope=col>card_present_flag</th><th scope=col>bpay_biller_code</th><th scope=col>account</th><th scope=col>currency</th><th scope=col>long_lat</th><th scope=col>txn_description</th><th scope=col>merchant_id</th><th scope=col>merchant_code</th><th scope=col>first_name</th><th scope=col>...</th><th scope=col>age</th><th scope=col>merchant_suburb</th><th scope=col>merchant_state</th><th scope=col>extraction</th><th scope=col>amount</th><th scope=col>transaction_id</th><th scope=col>country</th><th scope=col>customer_id</th><th scope=col>merchant_long_lat</th><th scope=col>movement</th></tr></thead>\n",
       "<tbody>\n",
       "\t<tr><td>authorized                          </td><td> 1                                  </td><td>NA                                  </td><td>ACC-1598451071                      </td><td>AUD                                 </td><td>153.41 -27.95                       </td><td>POS                                 </td><td>81c48296-73be-44a7-befa-d053f48ce7cd</td><td>NA                                  </td><td>Diana                               </td><td>...                                 </td><td>26                                  </td><td>Ashmore                             </td><td>QLD                                 </td><td>2018-08-01T01:01:15.000+0000        </td><td> 16.25                              </td><td>a623070bfead4541a6b0fff8a09e706c    </td><td>Australia                           </td><td>CUS-2487424745                      </td><td>153.38 -27.99                       </td><td>debit                               </td></tr>\n",
       "\t<tr><td>authorized                          </td><td> 0                                  </td><td>NA                                  </td><td>ACC-1598451071                      </td><td>AUD                                 </td><td>153.41 -27.95                       </td><td>SALES-POS                           </td><td>830a451c-316e-4a6a-bf25-e37caedca49e</td><td>NA                                  </td><td>Diana                               </td><td>...                                 </td><td>26                                  </td><td>Sydney                              </td><td>NSW                                 </td><td>2018-08-01T01:13:45.000+0000        </td><td> 14.19                              </td><td>13270a2a902145da9db4c951e04b51b9    </td><td>Australia                           </td><td>CUS-2487424745                      </td><td>151.21 -33.87                       </td><td>debit                               </td></tr>\n",
       "\t<tr><td>authorized                          </td><td> 1                                  </td><td>NA                                  </td><td>ACC-1222300524                      </td><td>AUD                                 </td><td>151.23 -33.94                       </td><td>POS                                 </td><td>835c231d-8cdf-4e96-859d-e9d571760cf0</td><td>NA                                  </td><td>Michael                             </td><td>...                                 </td><td>38                                  </td><td>Sydney                              </td><td>NSW                                 </td><td>2018-08-01T01:26:15.000+0000        </td><td>  6.42                              </td><td>feb79e7ecd7048a5a36ec889d1a94270    </td><td>Australia                           </td><td>CUS-2142601169                      </td><td>151.21 -33.87                       </td><td>debit                               </td></tr>\n",
       "\t<tr><td>authorized                          </td><td> 1                                  </td><td>NA                                  </td><td>ACC-1037050564                      </td><td>AUD                                 </td><td>153.10 -27.66                       </td><td>SALES-POS                           </td><td>48514682-c78a-4a88-b0da-2d6302e64673</td><td>NA                                  </td><td>Rhonda                              </td><td>...                                 </td><td>40                                  </td><td>Buderim                             </td><td>QLD                                 </td><td>2018-08-01T01:38:45.000+0000        </td><td> 40.90                              </td><td>2698170da3704fd981b15e64a006079e    </td><td>Australia                           </td><td>CUS-1614226872                      </td><td>153.05 -26.68                       </td><td>debit                               </td></tr>\n",
       "\t<tr><td>authorized                          </td><td> 1                                  </td><td>NA                                  </td><td>ACC-1598451071                      </td><td>AUD                                 </td><td>153.41 -27.95                       </td><td>SALES-POS                           </td><td>b4e02c10-0852-4273-b8fd-7b3395e32eb0</td><td>NA                                  </td><td>Diana                               </td><td>...                                 </td><td>26                                  </td><td>Mermaid Beach                       </td><td>QLD                                 </td><td>2018-08-01T01:51:15.000+0000        </td><td>  3.25                              </td><td>329adf79878c4cf0aeb4188b4691c266    </td><td>Australia                           </td><td>CUS-2487424745                      </td><td>153.44 -28.06                       </td><td>debit                               </td></tr>\n",
       "\t<tr><td>posted                              </td><td>NA                                  </td><td>NA                                  </td><td>ACC-1608363396                      </td><td>AUD                                 </td><td>151.22 -33.87                       </td><td>PAYMENT                             </td><td>NA                                  </td><td>NA                                  </td><td>Robert                              </td><td>...                                 </td><td>20                                  </td><td>NA                                  </td><td>NA                                  </td><td>2018-08-01T02:00:00.000+0000        </td><td>163.00                              </td><td>1005b48a6eda4ffd85e9b649dc9467d3    </td><td>Australia                           </td><td>CUS-2688605418                      </td><td>NA                                  </td><td>debit                               </td></tr>\n",
       "</tbody>\n",
       "</table>\n"
      ],
      "text/latex": [
       "\\begin{tabular}{r|lllllllllllllllllllllll}\n",
       " status & card\\_present\\_flag & bpay\\_biller\\_code & account & currency & long\\_lat & txn\\_description & merchant\\_id & merchant\\_code & first\\_name & ... & age & merchant\\_suburb & merchant\\_state & extraction & amount & transaction\\_id & country & customer\\_id & merchant\\_long\\_lat & movement\\\\\n",
       "\\hline\n",
       "\t authorized                           &  1                                   & NA                                   & ACC-1598451071                       & AUD                                  & 153.41 -27.95                        & POS                                  & 81c48296-73be-44a7-befa-d053f48ce7cd & NA                                   & Diana                                & ...                                  & 26                                   & Ashmore                              & QLD                                  & 2018-08-01T01:01:15.000+0000         &  16.25                               & a623070bfead4541a6b0fff8a09e706c     & Australia                            & CUS-2487424745                       & 153.38 -27.99                        & debit                               \\\\\n",
       "\t authorized                           &  0                                   & NA                                   & ACC-1598451071                       & AUD                                  & 153.41 -27.95                        & SALES-POS                            & 830a451c-316e-4a6a-bf25-e37caedca49e & NA                                   & Diana                                & ...                                  & 26                                   & Sydney                               & NSW                                  & 2018-08-01T01:13:45.000+0000         &  14.19                               & 13270a2a902145da9db4c951e04b51b9     & Australia                            & CUS-2487424745                       & 151.21 -33.87                        & debit                               \\\\\n",
       "\t authorized                           &  1                                   & NA                                   & ACC-1222300524                       & AUD                                  & 151.23 -33.94                        & POS                                  & 835c231d-8cdf-4e96-859d-e9d571760cf0 & NA                                   & Michael                              & ...                                  & 38                                   & Sydney                               & NSW                                  & 2018-08-01T01:26:15.000+0000         &   6.42                               & feb79e7ecd7048a5a36ec889d1a94270     & Australia                            & CUS-2142601169                       & 151.21 -33.87                        & debit                               \\\\\n",
       "\t authorized                           &  1                                   & NA                                   & ACC-1037050564                       & AUD                                  & 153.10 -27.66                        & SALES-POS                            & 48514682-c78a-4a88-b0da-2d6302e64673 & NA                                   & Rhonda                               & ...                                  & 40                                   & Buderim                              & QLD                                  & 2018-08-01T01:38:45.000+0000         &  40.90                               & 2698170da3704fd981b15e64a006079e     & Australia                            & CUS-1614226872                       & 153.05 -26.68                        & debit                               \\\\\n",
       "\t authorized                           &  1                                   & NA                                   & ACC-1598451071                       & AUD                                  & 153.41 -27.95                        & SALES-POS                            & b4e02c10-0852-4273-b8fd-7b3395e32eb0 & NA                                   & Diana                                & ...                                  & 26                                   & Mermaid Beach                        & QLD                                  & 2018-08-01T01:51:15.000+0000         &   3.25                               & 329adf79878c4cf0aeb4188b4691c266     & Australia                            & CUS-2487424745                       & 153.44 -28.06                        & debit                               \\\\\n",
       "\t posted                               & NA                                   & NA                                   & ACC-1608363396                       & AUD                                  & 151.22 -33.87                        & PAYMENT                              & NA                                   & NA                                   & Robert                               & ...                                  & 20                                   & NA                                   & NA                                   & 2018-08-01T02:00:00.000+0000         & 163.00                               & 1005b48a6eda4ffd85e9b649dc9467d3     & Australia                            & CUS-2688605418                       & NA                                   & debit                               \\\\\n",
       "\\end{tabular}\n"
      ],
      "text/markdown": [
       "\n",
       "| status | card_present_flag | bpay_biller_code | account | currency | long_lat | txn_description | merchant_id | merchant_code | first_name | ... | age | merchant_suburb | merchant_state | extraction | amount | transaction_id | country | customer_id | merchant_long_lat | movement |\n",
       "|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n",
       "| authorized                           |  1                                   | NA                                   | ACC-1598451071                       | AUD                                  | 153.41 -27.95                        | POS                                  | 81c48296-73be-44a7-befa-d053f48ce7cd | NA                                   | Diana                                | ...                                  | 26                                   | Ashmore                              | QLD                                  | 2018-08-01T01:01:15.000+0000         |  16.25                               | a623070bfead4541a6b0fff8a09e706c     | Australia                            | CUS-2487424745                       | 153.38 -27.99                        | debit                                |\n",
       "| authorized                           |  0                                   | NA                                   | ACC-1598451071                       | AUD                                  | 153.41 -27.95                        | SALES-POS                            | 830a451c-316e-4a6a-bf25-e37caedca49e | NA                                   | Diana                                | ...                                  | 26                                   | Sydney                               | NSW                                  | 2018-08-01T01:13:45.000+0000         |  14.19                               | 13270a2a902145da9db4c951e04b51b9     | Australia                            | CUS-2487424745                       | 151.21 -33.87                        | debit                                |\n",
       "| authorized                           |  1                                   | NA                                   | ACC-1222300524                       | AUD                                  | 151.23 -33.94                        | POS                                  | 835c231d-8cdf-4e96-859d-e9d571760cf0 | NA                                   | Michael                              | ...                                  | 38                                   | Sydney                               | NSW                                  | 2018-08-01T01:26:15.000+0000         |   6.42                               | feb79e7ecd7048a5a36ec889d1a94270     | Australia                            | CUS-2142601169                       | 151.21 -33.87                        | debit                                |\n",
       "| authorized                           |  1                                   | NA                                   | ACC-1037050564                       | AUD                                  | 153.10 -27.66                        | SALES-POS                            | 48514682-c78a-4a88-b0da-2d6302e64673 | NA                                   | Rhonda                               | ...                                  | 40                                   | Buderim                              | QLD                                  | 2018-08-01T01:38:45.000+0000         |  40.90                               | 2698170da3704fd981b15e64a006079e     | Australia                            | CUS-1614226872                       | 153.05 -26.68                        | debit                                |\n",
       "| authorized                           |  1                                   | NA                                   | ACC-1598451071                       | AUD                                  | 153.41 -27.95                        | SALES-POS                            | b4e02c10-0852-4273-b8fd-7b3395e32eb0 | NA                                   | Diana                                | ...                                  | 26                                   | Mermaid Beach                        | QLD                                  | 2018-08-01T01:51:15.000+0000         |   3.25                               | 329adf79878c4cf0aeb4188b4691c266     | Australia                            | CUS-2487424745                       | 153.44 -28.06                        | debit                                |\n",
       "| posted                               | NA                                   | NA                                   | ACC-1608363396                       | AUD                                  | 151.22 -33.87                        | PAYMENT                              | NA                                   | NA                                   | Robert                               | ...                                  | 20                                   | NA                                   | NA                                   | 2018-08-01T02:00:00.000+0000         | 163.00                               | 1005b48a6eda4ffd85e9b649dc9467d3     | Australia                            | CUS-2688605418                       | NA                                   | debit                                |\n",
       "\n"
      ],
      "text/plain": [
       "  status     card_present_flag bpay_biller_code account        currency\n",
       "1 authorized  1                NA               ACC-1598451071 AUD     \n",
       "2 authorized  0                NA               ACC-1598451071 AUD     \n",
       "3 authorized  1                NA               ACC-1222300524 AUD     \n",
       "4 authorized  1                NA               ACC-1037050564 AUD     \n",
       "5 authorized  1                NA               ACC-1598451071 AUD     \n",
       "6 posted     NA                NA               ACC-1608363396 AUD     \n",
       "  long_lat      txn_description merchant_id                         \n",
       "1 153.41 -27.95 POS             81c48296-73be-44a7-befa-d053f48ce7cd\n",
       "2 153.41 -27.95 SALES-POS       830a451c-316e-4a6a-bf25-e37caedca49e\n",
       "3 151.23 -33.94 POS             835c231d-8cdf-4e96-859d-e9d571760cf0\n",
       "4 153.10 -27.66 SALES-POS       48514682-c78a-4a88-b0da-2d6302e64673\n",
       "5 153.41 -27.95 SALES-POS       b4e02c10-0852-4273-b8fd-7b3395e32eb0\n",
       "6 151.22 -33.87 PAYMENT         NA                                  \n",
       "  merchant_code first_name ... age merchant_suburb merchant_state\n",
       "1 NA            Diana      ... 26  Ashmore         QLD           \n",
       "2 NA            Diana      ... 26  Sydney          NSW           \n",
       "3 NA            Michael    ... 38  Sydney          NSW           \n",
       "4 NA            Rhonda     ... 40  Buderim         QLD           \n",
       "5 NA            Diana      ... 26  Mermaid Beach   QLD           \n",
       "6 NA            Robert     ... 20  NA              NA            \n",
       "  extraction                   amount transaction_id                  \n",
       "1 2018-08-01T01:01:15.000+0000  16.25 a623070bfead4541a6b0fff8a09e706c\n",
       "2 2018-08-01T01:13:45.000+0000  14.19 13270a2a902145da9db4c951e04b51b9\n",
       "3 2018-08-01T01:26:15.000+0000   6.42 feb79e7ecd7048a5a36ec889d1a94270\n",
       "4 2018-08-01T01:38:45.000+0000  40.90 2698170da3704fd981b15e64a006079e\n",
       "5 2018-08-01T01:51:15.000+0000   3.25 329adf79878c4cf0aeb4188b4691c266\n",
       "6 2018-08-01T02:00:00.000+0000 163.00 1005b48a6eda4ffd85e9b649dc9467d3\n",
       "  country   customer_id    merchant_long_lat movement\n",
       "1 Australia CUS-2487424745 153.38 -27.99     debit   \n",
       "2 Australia CUS-2487424745 151.21 -33.87     debit   \n",
       "3 Australia CUS-2142601169 151.21 -33.87     debit   \n",
       "4 Australia CUS-1614226872 153.05 -26.68     debit   \n",
       "5 Australia CUS-2487424745 153.44 -28.06     debit   \n",
       "6 Australia CUS-2688605418 NA                debit   "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "setwd(\"C:/Users/Jason Chong/Documents/Forage ANZ/Data\")\n",
    "df = read_excel(\"transaction.xlsx\")\n",
    "head(df)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Data preparation "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "    status          card_present_flag bpay_biller_code   account         \n",
       " Length:12043       Min.   :0.000     Min.   :0        Length:12043      \n",
       " Class :character   1st Qu.:1.000     1st Qu.:0        Class :character  \n",
       " Mode  :character   Median :1.000     Median :0        Mode  :character  \n",
       "                    Mean   :0.803     Mean   :0                          \n",
       "                    3rd Qu.:1.000     3rd Qu.:0                          \n",
       "                    Max.   :1.000     Max.   :0                          \n",
       "                    NA's   :4326      NA's   :11160                      \n",
       "   currency           long_lat         txn_description    merchant_id       \n",
       " Length:12043       Length:12043       Length:12043       Length:12043      \n",
       " Class :character   Class :character   Class :character   Class :character  \n",
       " Mode  :character   Mode  :character   Mode  :character   Mode  :character  \n",
       "                                                                            \n",
       "                                                                            \n",
       "                                                                            \n",
       "                                                                            \n",
       " merchant_code    first_name           balance         \n",
       " Min.   :0       Length:12043       Min.   :     0.24  \n",
       " 1st Qu.:0       Class :character   1st Qu.:  3158.59  \n",
       " Median :0       Mode  :character   Median :  6432.01  \n",
       " Mean   :0                          Mean   : 14704.20  \n",
       " 3rd Qu.:0                          3rd Qu.: 12465.94  \n",
       " Max.   :0                          Max.   :267128.52  \n",
       " NA's   :11160                                         \n",
       "      date                        gender               age       \n",
       " Min.   :2018-08-01 00:00:00   Length:12043       Min.   :18.00  \n",
       " 1st Qu.:2018-08-24 00:00:00   Class :character   1st Qu.:22.00  \n",
       " Median :2018-09-16 00:00:00   Mode  :character   Median :28.00  \n",
       " Mean   :2018-09-15 21:27:39                      Mean   :30.58  \n",
       " 3rd Qu.:2018-10-09 00:00:00                      3rd Qu.:38.00  \n",
       " Max.   :2018-10-31 00:00:00                      Max.   :78.00  \n",
       "                                                                 \n",
       " merchant_suburb    merchant_state      extraction            amount       \n",
       " Length:12043       Length:12043       Length:12043       Min.   :   0.10  \n",
       " Class :character   Class :character   Class :character   1st Qu.:  16.00  \n",
       " Mode  :character   Mode  :character   Mode  :character   Median :  29.00  \n",
       "                                                          Mean   : 187.93  \n",
       "                                                          3rd Qu.:  53.66  \n",
       "                                                          Max.   :8835.98  \n",
       "                                                                           \n",
       " transaction_id       country          customer_id        merchant_long_lat \n",
       " Length:12043       Length:12043       Length:12043       Length:12043      \n",
       " Class :character   Class :character   Class :character   Class :character  \n",
       " Mode  :character   Mode  :character   Mode  :character   Mode  :character  \n",
       "                                                                            \n",
       "                                                                            \n",
       "                                                                            \n",
       "                                                                            \n",
       "   movement        \n",
       " Length:12043      \n",
       " Class :character  \n",
       " Mode  :character  \n",
       "                   \n",
       "                   \n",
       "                   \n",
       "                   "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "summary(df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Classes 'tbl_df', 'tbl' and 'data.frame':\t12043 obs. of  23 variables:\n",
      " $ status           : chr  \"authorized\" \"authorized\" \"authorized\" \"authorized\" ...\n",
      " $ card_present_flag: num  1 0 1 1 1 NA 1 1 1 NA ...\n",
      " $ bpay_biller_code : num  NA NA NA NA NA NA NA NA NA NA ...\n",
      " $ account          : chr  \"ACC-1598451071\" \"ACC-1598451071\" \"ACC-1222300524\" \"ACC-1037050564\" ...\n",
      " $ currency         : chr  \"AUD\" \"AUD\" \"AUD\" \"AUD\" ...\n",
      " $ long_lat         : chr  \"153.41 -27.95\" \"153.41 -27.95\" \"151.23 -33.94\" \"153.10 -27.66\" ...\n",
      " $ txn_description  : chr  \"POS\" \"SALES-POS\" \"POS\" \"SALES-POS\" ...\n",
      " $ merchant_id      : chr  \"81c48296-73be-44a7-befa-d053f48ce7cd\" \"830a451c-316e-4a6a-bf25-e37caedca49e\" \"835c231d-8cdf-4e96-859d-e9d571760cf0\" \"48514682-c78a-4a88-b0da-2d6302e64673\" ...\n",
      " $ merchant_code    : num  NA NA NA NA NA NA NA NA NA NA ...\n",
      " $ first_name       : chr  \"Diana\" \"Diana\" \"Michael\" \"Rhonda\" ...\n",
      " $ balance          : num  35.39 21.2 5.71 2117.22 17.95 ...\n",
      " $ date             : POSIXct, format: \"2018-08-01\" \"2018-08-01\" ...\n",
      " $ gender           : chr  \"F\" \"F\" \"M\" \"F\" ...\n",
      " $ age              : num  26 26 38 40 26 20 43 43 27 40 ...\n",
      " $ merchant_suburb  : chr  \"Ashmore\" \"Sydney\" \"Sydney\" \"Buderim\" ...\n",
      " $ merchant_state   : chr  \"QLD\" \"NSW\" \"NSW\" \"QLD\" ...\n",
      " $ extraction       : chr  \"2018-08-01T01:01:15.000+0000\" \"2018-08-01T01:13:45.000+0000\" \"2018-08-01T01:26:15.000+0000\" \"2018-08-01T01:38:45.000+0000\" ...\n",
      " $ amount           : num  16.25 14.19 6.42 40.9 3.25 ...\n",
      " $ transaction_id   : chr  \"a623070bfead4541a6b0fff8a09e706c\" \"13270a2a902145da9db4c951e04b51b9\" \"feb79e7ecd7048a5a36ec889d1a94270\" \"2698170da3704fd981b15e64a006079e\" ...\n",
      " $ country          : chr  \"Australia\" \"Australia\" \"Australia\" \"Australia\" ...\n",
      " $ customer_id      : chr  \"CUS-2487424745\" \"CUS-2487424745\" \"CUS-2142601169\" \"CUS-1614226872\" ...\n",
      " $ merchant_long_lat: chr  \"153.38 -27.99\" \"151.21 -33.87\" \"151.21 -33.87\" \"153.05 -26.68\" ...\n",
      " $ movement         : chr  \"debit\" \"debit\" \"debit\" \"debit\" ...\n"
     ]
    }
   ],
   "source": [
    "str(df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Change the format of the date column \n",
    "df$date = as.Date(df$date, format = \"%d/%m/%Y\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<time datetime=\"2018-08-16\">2018-08-16</time>"
      ],
      "text/latex": [
       "2018-08-16"
      ],
      "text/markdown": [
       "2018-08-16"
      ],
      "text/plain": [
       "[1] \"2018-08-16\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# The dataset only contain records for 91 days, one day is missing \n",
    "date_range = seq(min(df$date), max(df$date), by = 1)\n",
    "date_range[!date_range %in% df$date]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Derive weekday and hour data of each transaction \n",
    "df$extraction = as.character(df$extraction)\n",
    "df$hour = hour(as.POSIXct(substr(df$extraction,12,19), format = \"%H:%M:%S\"))\n",
    "df$weekday = weekdays(df$date)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "100"
      ],
      "text/latex": [
       "100"
      ],
      "text/markdown": [
       "100"
      ],
      "text/plain": [
       "[1] 100"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Confirm the one-to-one link of account_id and customer_id\n",
    "df %>% select(account, customer_id) %>% unique() %>% nrow()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Split customer and merchant long_lat into individual columns for analysis\n",
    "dfloc = df[, c(\"long_lat\", \"merchant_long_lat\")]\n",
    "dfloc = dfloc %>% separate(\"long_lat\", c(\"c_long\", \"c_lat\"), sep = \" \")\n",
    "dfloc = dfloc %>% separate(\"merchant_long_lat\", c(\"m_long\", \"m_lat\"), sep = \" \")\n",
    "dfloc = data.frame(sapply(dfloc, as.numeric))\n",
    "df = cbind(df, dfloc)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table>\n",
       "<thead><tr><th scope=col>long_lat</th><th scope=col>c_long</th><th scope=col>c_lat</th><th scope=col>merchant_long_lat</th><th scope=col>m_long</th><th scope=col>m_lat</th></tr></thead>\n",
       "<tbody>\n",
       "\t<tr><td>153.41 -27.95</td><td>153.41       </td><td>-27.95       </td><td>153.38 -27.99</td><td>153.38       </td><td>-27.99       </td></tr>\n",
       "\t<tr><td>153.41 -27.95</td><td>153.41       </td><td>-27.95       </td><td>151.21 -33.87</td><td>151.21       </td><td>-33.87       </td></tr>\n",
       "\t<tr><td>151.23 -33.94</td><td>151.23       </td><td>-33.94       </td><td>151.21 -33.87</td><td>151.21       </td><td>-33.87       </td></tr>\n",
       "\t<tr><td>153.10 -27.66</td><td>153.10       </td><td>-27.66       </td><td>153.05 -26.68</td><td>153.05       </td><td>-26.68       </td></tr>\n",
       "\t<tr><td>153.41 -27.95</td><td>153.41       </td><td>-27.95       </td><td>153.44 -28.06</td><td>153.44       </td><td>-28.06       </td></tr>\n",
       "\t<tr><td>151.22 -33.87</td><td>151.22       </td><td>-33.87       </td><td>NA           </td><td>    NA       </td><td>    NA       </td></tr>\n",
       "</tbody>\n",
       "</table>\n"
      ],
      "text/latex": [
       "\\begin{tabular}{r|llllll}\n",
       " long\\_lat & c\\_long & c\\_lat & merchant\\_long\\_lat & m\\_long & m\\_lat\\\\\n",
       "\\hline\n",
       "\t 153.41 -27.95 & 153.41        & -27.95        & 153.38 -27.99 & 153.38        & -27.99       \\\\\n",
       "\t 153.41 -27.95 & 153.41        & -27.95        & 151.21 -33.87 & 151.21        & -33.87       \\\\\n",
       "\t 151.23 -33.94 & 151.23        & -33.94        & 151.21 -33.87 & 151.21        & -33.87       \\\\\n",
       "\t 153.10 -27.66 & 153.10        & -27.66        & 153.05 -26.68 & 153.05        & -26.68       \\\\\n",
       "\t 153.41 -27.95 & 153.41        & -27.95        & 153.44 -28.06 & 153.44        & -28.06       \\\\\n",
       "\t 151.22 -33.87 & 151.22        & -33.87        & NA            &     NA        &     NA       \\\\\n",
       "\\end{tabular}\n"
      ],
      "text/markdown": [
       "\n",
       "| long_lat | c_long | c_lat | merchant_long_lat | m_long | m_lat |\n",
       "|---|---|---|---|---|---|\n",
       "| 153.41 -27.95 | 153.41        | -27.95        | 153.38 -27.99 | 153.38        | -27.99        |\n",
       "| 153.41 -27.95 | 153.41        | -27.95        | 151.21 -33.87 | 151.21        | -33.87        |\n",
       "| 151.23 -33.94 | 151.23        | -33.94        | 151.21 -33.87 | 151.21        | -33.87        |\n",
       "| 153.10 -27.66 | 153.10        | -27.66        | 153.05 -26.68 | 153.05        | -26.68        |\n",
       "| 153.41 -27.95 | 153.41        | -27.95        | 153.44 -28.06 | 153.44        | -28.06        |\n",
       "| 151.22 -33.87 | 151.22        | -33.87        | NA            |     NA        |     NA        |\n",
       "\n"
      ],
      "text/plain": [
       "  long_lat      c_long c_lat  merchant_long_lat m_long m_lat \n",
       "1 153.41 -27.95 153.41 -27.95 153.38 -27.99     153.38 -27.99\n",
       "2 153.41 -27.95 153.41 -27.95 151.21 -33.87     151.21 -33.87\n",
       "3 151.23 -33.94 151.23 -33.94 151.21 -33.87     151.21 -33.87\n",
       "4 153.10 -27.66 153.10 -27.66 153.05 -26.68     153.05 -26.68\n",
       "5 153.41 -27.95 153.41 -27.95 153.44 -28.06     153.44 -28.06\n",
       "6 151.22 -33.87 151.22 -33.87 NA                    NA     NA"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "head(df[, c(\"long_lat\", \"c_long\", \"c_lat\", \"merchant_long_lat\", \"m_long\", \"m_lat\")])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "1"
      ],
      "text/latex": [
       "1"
      ],
      "text/markdown": [
       "1"
      ],
      "text/plain": [
       "[1] 1"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Check the range of customer location\n",
    "# Filtering out transactions for those who do not reside in Australia\n",
    "df_temp = df %>% filter(!(c_long > 113 & c_long < 154 & c_lat > (-44) & c_lat < (-10)))\n",
    "length(unique(df_temp$customer_id))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Location information suggests there is one customer who resides outside Australia. However, all of his/her transaction histories occured within Australia, therefore these records are included for further analysis. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<dl class=dl-horizontal>\n",
       "\t<dt>status</dt>\n",
       "\t\t<dd>0</dd>\n",
       "\t<dt>card_present_flag</dt>\n",
       "\t\t<dd>4326</dd>\n",
       "\t<dt>bpay_biller_code</dt>\n",
       "\t\t<dd>11160</dd>\n",
       "\t<dt>account</dt>\n",
       "\t\t<dd>0</dd>\n",
       "\t<dt>currency</dt>\n",
       "\t\t<dd>0</dd>\n",
       "\t<dt>long_lat</dt>\n",
       "\t\t<dd>0</dd>\n",
       "\t<dt>txn_description</dt>\n",
       "\t\t<dd>0</dd>\n",
       "\t<dt>merchant_id</dt>\n",
       "\t\t<dd>4326</dd>\n",
       "\t<dt>merchant_code</dt>\n",
       "\t\t<dd>11160</dd>\n",
       "\t<dt>first_name</dt>\n",
       "\t\t<dd>0</dd>\n",
       "\t<dt>balance</dt>\n",
       "\t\t<dd>0</dd>\n",
       "\t<dt>date</dt>\n",
       "\t\t<dd>0</dd>\n",
       "\t<dt>gender</dt>\n",
       "\t\t<dd>0</dd>\n",
       "\t<dt>age</dt>\n",
       "\t\t<dd>0</dd>\n",
       "\t<dt>merchant_suburb</dt>\n",
       "\t\t<dd>4326</dd>\n",
       "\t<dt>merchant_state</dt>\n",
       "\t\t<dd>4326</dd>\n",
       "\t<dt>extraction</dt>\n",
       "\t\t<dd>0</dd>\n",
       "\t<dt>amount</dt>\n",
       "\t\t<dd>0</dd>\n",
       "\t<dt>transaction_id</dt>\n",
       "\t\t<dd>0</dd>\n",
       "\t<dt>country</dt>\n",
       "\t\t<dd>0</dd>\n",
       "\t<dt>customer_id</dt>\n",
       "\t\t<dd>0</dd>\n",
       "\t<dt>merchant_long_lat</dt>\n",
       "\t\t<dd>4326</dd>\n",
       "\t<dt>movement</dt>\n",
       "\t\t<dd>0</dd>\n",
       "\t<dt>hour</dt>\n",
       "\t\t<dd>0</dd>\n",
       "\t<dt>weekday</dt>\n",
       "\t\t<dd>0</dd>\n",
       "\t<dt>c_long</dt>\n",
       "\t\t<dd>0</dd>\n",
       "\t<dt>c_lat</dt>\n",
       "\t\t<dd>0</dd>\n",
       "\t<dt>m_long</dt>\n",
       "\t\t<dd>4326</dd>\n",
       "\t<dt>m_lat</dt>\n",
       "\t\t<dd>4326</dd>\n",
       "</dl>\n"
      ],
      "text/latex": [
       "\\begin{description*}\n",
       "\\item[status] 0\n",
       "\\item[card\\textbackslash{}\\_present\\textbackslash{}\\_flag] 4326\n",
       "\\item[bpay\\textbackslash{}\\_biller\\textbackslash{}\\_code] 11160\n",
       "\\item[account] 0\n",
       "\\item[currency] 0\n",
       "\\item[long\\textbackslash{}\\_lat] 0\n",
       "\\item[txn\\textbackslash{}\\_description] 0\n",
       "\\item[merchant\\textbackslash{}\\_id] 4326\n",
       "\\item[merchant\\textbackslash{}\\_code] 11160\n",
       "\\item[first\\textbackslash{}\\_name] 0\n",
       "\\item[balance] 0\n",
       "\\item[date] 0\n",
       "\\item[gender] 0\n",
       "\\item[age] 0\n",
       "\\item[merchant\\textbackslash{}\\_suburb] 4326\n",
       "\\item[merchant\\textbackslash{}\\_state] 4326\n",
       "\\item[extraction] 0\n",
       "\\item[amount] 0\n",
       "\\item[transaction\\textbackslash{}\\_id] 0\n",
       "\\item[country] 0\n",
       "\\item[customer\\textbackslash{}\\_id] 0\n",
       "\\item[merchant\\textbackslash{}\\_long\\textbackslash{}\\_lat] 4326\n",
       "\\item[movement] 0\n",
       "\\item[hour] 0\n",
       "\\item[weekday] 0\n",
       "\\item[c\\textbackslash{}\\_long] 0\n",
       "\\item[c\\textbackslash{}\\_lat] 0\n",
       "\\item[m\\textbackslash{}\\_long] 4326\n",
       "\\item[m\\textbackslash{}\\_lat] 4326\n",
       "\\end{description*}\n"
      ],
      "text/markdown": [
       "status\n",
       ":   0card_present_flag\n",
       ":   4326bpay_biller_code\n",
       ":   11160account\n",
       ":   0currency\n",
       ":   0long_lat\n",
       ":   0txn_description\n",
       ":   0merchant_id\n",
       ":   4326merchant_code\n",
       ":   11160first_name\n",
       ":   0balance\n",
       ":   0date\n",
       ":   0gender\n",
       ":   0age\n",
       ":   0merchant_suburb\n",
       ":   4326merchant_state\n",
       ":   4326extraction\n",
       ":   0amount\n",
       ":   0transaction_id\n",
       ":   0country\n",
       ":   0customer_id\n",
       ":   0merchant_long_lat\n",
       ":   4326movement\n",
       ":   0hour\n",
       ":   0weekday\n",
       ":   0c_long\n",
       ":   0c_lat\n",
       ":   0m_long\n",
       ":   4326m_lat\n",
       ":   4326\n",
       "\n"
      ],
      "text/plain": [
       "           status card_present_flag  bpay_biller_code           account \n",
       "                0              4326             11160                 0 \n",
       "         currency          long_lat   txn_description       merchant_id \n",
       "                0                 0                 0              4326 \n",
       "    merchant_code        first_name           balance              date \n",
       "            11160                 0                 0                 0 \n",
       "           gender               age   merchant_suburb    merchant_state \n",
       "                0                 0              4326              4326 \n",
       "       extraction            amount    transaction_id           country \n",
       "                0                 0                 0                 0 \n",
       "      customer_id merchant_long_lat          movement              hour \n",
       "                0              4326                 0                 0 \n",
       "          weekday            c_long             c_lat            m_long \n",
       "                0                 0                 0              4326 \n",
       "            m_lat \n",
       "             4326 "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<dl class=dl-horizontal>\n",
       "\t<dt>status</dt>\n",
       "\t\t<dd>2</dd>\n",
       "\t<dt>card_present_flag</dt>\n",
       "\t\t<dd>3</dd>\n",
       "\t<dt>bpay_biller_code</dt>\n",
       "\t\t<dd>2</dd>\n",
       "\t<dt>account</dt>\n",
       "\t\t<dd>100</dd>\n",
       "\t<dt>currency</dt>\n",
       "\t\t<dd>1</dd>\n",
       "\t<dt>long_lat</dt>\n",
       "\t\t<dd>100</dd>\n",
       "\t<dt>txn_description</dt>\n",
       "\t\t<dd>6</dd>\n",
       "\t<dt>merchant_id</dt>\n",
       "\t\t<dd>5726</dd>\n",
       "\t<dt>merchant_code</dt>\n",
       "\t\t<dd>2</dd>\n",
       "\t<dt>first_name</dt>\n",
       "\t\t<dd>80</dd>\n",
       "\t<dt>balance</dt>\n",
       "\t\t<dd>12006</dd>\n",
       "\t<dt>date</dt>\n",
       "\t\t<dd>91</dd>\n",
       "\t<dt>gender</dt>\n",
       "\t\t<dd>2</dd>\n",
       "\t<dt>age</dt>\n",
       "\t\t<dd>33</dd>\n",
       "\t<dt>merchant_suburb</dt>\n",
       "\t\t<dd>1610</dd>\n",
       "\t<dt>merchant_state</dt>\n",
       "\t\t<dd>9</dd>\n",
       "\t<dt>extraction</dt>\n",
       "\t\t<dd>9442</dd>\n",
       "\t<dt>amount</dt>\n",
       "\t\t<dd>4457</dd>\n",
       "\t<dt>transaction_id</dt>\n",
       "\t\t<dd>12043</dd>\n",
       "\t<dt>country</dt>\n",
       "\t\t<dd>1</dd>\n",
       "\t<dt>customer_id</dt>\n",
       "\t\t<dd>100</dd>\n",
       "\t<dt>merchant_long_lat</dt>\n",
       "\t\t<dd>2704</dd>\n",
       "\t<dt>movement</dt>\n",
       "\t\t<dd>2</dd>\n",
       "\t<dt>hour</dt>\n",
       "\t\t<dd>24</dd>\n",
       "\t<dt>weekday</dt>\n",
       "\t\t<dd>7</dd>\n",
       "\t<dt>c_long</dt>\n",
       "\t\t<dd>87</dd>\n",
       "\t<dt>c_lat</dt>\n",
       "\t\t<dd>85</dd>\n",
       "\t<dt>m_long</dt>\n",
       "\t\t<dd>719</dd>\n",
       "\t<dt>m_lat</dt>\n",
       "\t\t<dd>670</dd>\n",
       "</dl>\n"
      ],
      "text/latex": [
       "\\begin{description*}\n",
       "\\item[status] 2\n",
       "\\item[card\\textbackslash{}\\_present\\textbackslash{}\\_flag] 3\n",
       "\\item[bpay\\textbackslash{}\\_biller\\textbackslash{}\\_code] 2\n",
       "\\item[account] 100\n",
       "\\item[currency] 1\n",
       "\\item[long\\textbackslash{}\\_lat] 100\n",
       "\\item[txn\\textbackslash{}\\_description] 6\n",
       "\\item[merchant\\textbackslash{}\\_id] 5726\n",
       "\\item[merchant\\textbackslash{}\\_code] 2\n",
       "\\item[first\\textbackslash{}\\_name] 80\n",
       "\\item[balance] 12006\n",
       "\\item[date] 91\n",
       "\\item[gender] 2\n",
       "\\item[age] 33\n",
       "\\item[merchant\\textbackslash{}\\_suburb] 1610\n",
       "\\item[merchant\\textbackslash{}\\_state] 9\n",
       "\\item[extraction] 9442\n",
       "\\item[amount] 4457\n",
       "\\item[transaction\\textbackslash{}\\_id] 12043\n",
       "\\item[country] 1\n",
       "\\item[customer\\textbackslash{}\\_id] 100\n",
       "\\item[merchant\\textbackslash{}\\_long\\textbackslash{}\\_lat] 2704\n",
       "\\item[movement] 2\n",
       "\\item[hour] 24\n",
       "\\item[weekday] 7\n",
       "\\item[c\\textbackslash{}\\_long] 87\n",
       "\\item[c\\textbackslash{}\\_lat] 85\n",
       "\\item[m\\textbackslash{}\\_long] 719\n",
       "\\item[m\\textbackslash{}\\_lat] 670\n",
       "\\end{description*}\n"
      ],
      "text/markdown": [
       "status\n",
       ":   2card_present_flag\n",
       ":   3bpay_biller_code\n",
       ":   2account\n",
       ":   100currency\n",
       ":   1long_lat\n",
       ":   100txn_description\n",
       ":   6merchant_id\n",
       ":   5726merchant_code\n",
       ":   2first_name\n",
       ":   80balance\n",
       ":   12006date\n",
       ":   91gender\n",
       ":   2age\n",
       ":   33merchant_suburb\n",
       ":   1610merchant_state\n",
       ":   9extraction\n",
       ":   9442amount\n",
       ":   4457transaction_id\n",
       ":   12043country\n",
       ":   1customer_id\n",
       ":   100merchant_long_lat\n",
       ":   2704movement\n",
       ":   2hour\n",
       ":   24weekday\n",
       ":   7c_long\n",
       ":   87c_lat\n",
       ":   85m_long\n",
       ":   719m_lat\n",
       ":   670\n",
       "\n"
      ],
      "text/plain": [
       "           status card_present_flag  bpay_biller_code           account \n",
       "                2                 3                 2               100 \n",
       "         currency          long_lat   txn_description       merchant_id \n",
       "                1               100                 6              5726 \n",
       "    merchant_code        first_name           balance              date \n",
       "                2                80             12006                91 \n",
       "           gender               age   merchant_suburb    merchant_state \n",
       "                2                33              1610                 9 \n",
       "       extraction            amount    transaction_id           country \n",
       "             9442              4457             12043                 1 \n",
       "      customer_id merchant_long_lat          movement              hour \n",
       "              100              2704                 2                24 \n",
       "          weekday            c_long             c_lat            m_long \n",
       "                7                87                85               719 \n",
       "            m_lat \n",
       "              670 "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Check the distribution of missing values\n",
    "apply(df, 2, function(x) sum(is.na(x) | x == \"\"))\n",
    "\n",
    "# Check the number of unique values for each column\n",
    "apply(df, 2, function(x) length(unique(x)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Gather some interesting overall insights about the data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "    status          card_present_flag bpay_biller_code   account         \n",
       " Length:7717        Min.   :0.0000    Min.   : NA      Length:7717       \n",
       " Class :character   1st Qu.:1.0000    1st Qu.: NA      Class :character  \n",
       " Mode  :character   Median :1.0000    Median : NA      Mode  :character  \n",
       "                    Mean   :0.8026    Mean   :NaN                        \n",
       "                    3rd Qu.:1.0000    3rd Qu.: NA                        \n",
       "                    Max.   :1.0000    Max.   : NA                        \n",
       "                                      NA's   :7717                       \n",
       "   currency           long_lat         txn_description    merchant_id       \n",
       " Length:7717        Length:7717        Length:7717        Length:7717       \n",
       " Class :character   Class :character   Class :character   Class :character  \n",
       " Mode  :character   Mode  :character   Mode  :character   Mode  :character  \n",
       "                                                                            \n",
       "                                                                            \n",
       "                                                                            \n",
       "                                                                            \n",
       " merchant_code   first_name           balance               date           \n",
       " Min.   : NA    Length:7717        Min.   :     0.24   Min.   :2018-08-01  \n",
       " 1st Qu.: NA    Class :character   1st Qu.:  2723.89   1st Qu.:2018-08-24  \n",
       " Median : NA    Mode  :character   Median :  5286.04   Median :2018-09-16  \n",
       " Mean   :NaN                       Mean   : 11871.06   Mean   :2018-09-15  \n",
       " 3rd Qu.: NA                       3rd Qu.: 10278.29   3rd Qu.:2018-10-08  \n",
       " Max.   : NA                       Max.   :267093.66   Max.   :2018-10-31  \n",
       " NA's   :7717                                                              \n",
       "    gender               age        merchant_suburb    merchant_state    \n",
       " Length:7717        Min.   :18.00   Length:7717        Length:7717       \n",
       " Class :character   1st Qu.:23.00   Class :character   Class :character  \n",
       " Mode  :character   Median :27.00   Mode  :character   Mode  :character  \n",
       "                    Mean   :30.05                                        \n",
       "                    3rd Qu.:38.00                                        \n",
       "                    Max.   :78.00                                        \n",
       "                                                                         \n",
       "  extraction            amount        transaction_id       country         \n",
       " Length:7717        Min.   :   0.10   Length:7717        Length:7717       \n",
       " Class :character   1st Qu.:  12.08   Class :character   Class :character  \n",
       " Mode  :character   Median :  19.70   Mode  :character   Mode  :character  \n",
       "                    Mean   :  40.15                                        \n",
       "                    3rd Qu.:  33.91                                        \n",
       "                    Max.   :7081.09                                        \n",
       "                                                                           \n",
       " customer_id        merchant_long_lat    movement              hour      \n",
       " Length:7717        Length:7717        Length:7717        Min.   : 0.00  \n",
       " Class :character   Class :character   Class :character   1st Qu.: 9.00  \n",
       " Mode  :character   Mode  :character   Mode  :character   Median :14.00  \n",
       "                                                          Mean   :13.41  \n",
       "                                                          3rd Qu.:19.00  \n",
       "                                                          Max.   :23.00  \n",
       "                                                                         \n",
       "   weekday              c_long          c_lat             m_long     \n",
       " Length:7717        Min.   :114.6   Min.   :-573.00   Min.   :113.8  \n",
       " Class :character   1st Qu.:138.7   1st Qu.: -37.58   1st Qu.:144.7  \n",
       " Mode  :character   Median :145.7   Median : -33.78   Median :145.8  \n",
       "                    Mean   :143.8   Mean   : -37.68   Mean   :143.4  \n",
       "                    3rd Qu.:151.2   3rd Qu.: -27.95   3rd Qu.:151.2  \n",
       "                    Max.   :255.0   Max.   : -12.37   Max.   :153.6  \n",
       "                                                                     \n",
       "     m_lat       \n",
       " Min.   :-43.31  \n",
       " 1st Qu.:-37.71  \n",
       " Median :-33.84  \n",
       " Mean   :-32.75  \n",
       " 3rd Qu.:-29.44  \n",
       " Max.   :-12.33  \n",
       "                 "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Filter out purchase transactions only\n",
    "# Assuming purchase transactions must be associated with a merchant (have a merchant id)\n",
    "df_temp = df %>% filter(merchant_id != \"\")\n",
    "\n",
    "# It turned out that is equivalent to excluding following categories of transactions\n",
    "df_csmp = df %>% filter(!(txn_description %in% c(\"PAY/SALARY\", \"INTER BANK\", \"PHONE BANK\", \"PAYMENT\")))\n",
    "\n",
    "summary(df_csmp)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA0gAAANICAMAAADKOT/pAAAAMFBMVEUAAABNTU1oaGh8fHyM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",
      "text/plain": [
       "Plot with title \"Histogram of purchase transaction amount\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Visualise the distribution of transaction amount\n",
    "hist(df_csmp$amount[!df_csmp$amount %in% boxplot.stats(df_csmp$amount)$out], xlab = \"Transaction amount\", main = \"Histogram of purchase transaction amount\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA0gAAANICAMAAADKOT/pAAAAMFBMVEUAAABNTU1oaGh8fHyM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\nAAEECRBAkAAB/5vK8VdPklSuAAAAAElFTkSuQmCC",
      "text/plain": [
       "Plot with title \"Histogram of overall transaction amount\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "hist(df$amount[!df$amount %in% boxplot.stats(df$amount)$out], xlab = \"Transaction amount\", main = \"Histogram of overall transaction amount\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA0gAAANICAMAAADKOT/pAAAAMFBMVEUAAABNTU1oaGh8fHyM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\nTkSuQmCC",
      "text/plain": [
       "Plot with title \"Histogram of customers' monthly transaction volume\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Visualise customers' average monthly transaction volume\n",
    "df2 = df %>% group_by(customer_id) %>% summarise(mon_avg_vol = round(n()/3, 0))\n",
    "hist(df2$mon_avg_vol, xlab = \"Monthly transaction volume\", ylab = \"Number of customers\", main = \"Histogram of customers' monthly transaction volume\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA0gAAANICAMAAADKOT/pAAAANlBMVEUAAAAzMzNNTU1oaGh8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\nTkSuQmCC",
      "text/plain": [
       "plot without title"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Visualise transaction volume over an average week\n",
    "\n",
    "df3 = df %>% select(date, weekday) %>% group_by(date, weekday) %>% summarise(daily_avg_vol = n()) %>% group_by(weekday) %>% summarise(avg_vol = mean(daily_avg_vol, na.rm = TRUE))\n",
    "df3$weekday = factor(df3$weekday, levels = c(\"Monday\", \"Tuesday\", \"Wednesday\", \"Thursday\", \"Friday\", \"Saturday\", \"Sunday\"))\n",
    "ggplot(df3, aes(x = weekday, y = avg_vol)) + geom_point() + geom_line(aes(group = 1)) + ggtitle(\"Average transaction volume by weekday\") + labs(x = \"Weekday\", y = \"Transaction volume\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA0gAAANICAMAAADKOT/pAAAANlBMVEUAAAAzMzNNTU1oaGh8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\nEiEKQSRCFIJIhCgEkQhRCCIRopD/AQbvk8QOFnozAAAAAElFTkSuQmCC",
      "text/plain": [
       "plot without title"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Visualise transaction volume over an average week.\n",
    "df4 <- df %>% select(date,hour) %>% group_by(date,hour) %>% summarize(trans_vol=n()) %>% group_by(hour) %>% summarize(trans_vol_per_hr = mean(trans_vol,na.rm=TRUE))\n",
    "\n",
    "ggplot(df4, aes(x = hour, y=trans_vol_per_hr)) + geom_point() + geom_line(aes(group = 1)) + ggtitle(\"Average transaction volume by hour\") + labs(x = \"Hour\", y = \"Transaction volume\") + expand_limits(y = 0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Check the salary payment frequency of each customer\n",
    "df_inc = data.frame(customer_id = unique(df_csmp$customer_id))\n",
    "\n",
    "# Create a mode function that will be used to find out what is the salary payment frequency\n",
    "# R does not have a standard built-in function to calculate mode \n",
    "Mode = function(x){\n",
    "    ux = unique(x)\n",
    "    ux[which.max(tabulate(match(x, ux)))]\n",
    "}\n",
    "\n",
    "# Loop through all salary payments for each customer\n",
    "# Assume the salary level is constant for each customer over the observed period\n",
    "for (i in seq(nrow(df_inc))){\n",
    "    trans_data = df[df$customer_id == df_inc$customer_id[i] & df$txn_description == \"PAY/SALARY\", c(\"amount\", \"date\")] %>% group_by(date) %>% summarise(amount = sum(amount))\n",
    "    total_s = sum(trans_data$amount)\n",
    "    count = dim(trans_data)[1]\n",
    "    if (count == 0){\n",
    "        df_inc$freq[i] = NA\n",
    "        df_inc$level[i] = NA\n",
    "    } else {\n",
    "        s = c()\n",
    "        lvl = c()\n",
    "        for (j in seq(count - 1)){\n",
    "            s = c(s, (trans_data$date[j+1] - trans_data$date[j]))\n",
    "            lvl = c(lvl, trans_data$amount[j]) \n",
    "        }\n",
    "        lvl = c(lvl, tail(trans_data$amount, n = 1))\n",
    "        df_inc$freq[i] = Mode(s)\n",
    "        df_inc$level[i] = Mode(lvl)\n",
    "    }\n",
    "}\n",
    "\n",
    "df_inc$annual_salary = df_inc$level / df_inc$freq * 365.25"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table>\n",
       "<thead><tr><th scope=col>customer_id</th><th scope=col>freq</th><th scope=col>level</th><th scope=col>annual_salary</th></tr></thead>\n",
       "<tbody>\n",
       "\t<tr><td>CUS-2487424745</td><td> 7            </td><td>1013.67       </td><td>52891.85      </td></tr>\n",
       "\t<tr><td>CUS-2142601169</td><td> 7            </td><td>1002.13       </td><td>52289.71      </td></tr>\n",
       "\t<tr><td>CUS-1614226872</td><td> 7            </td><td> 892.09       </td><td>46547.98      </td></tr>\n",
       "\t<tr><td>CUS-4123612273</td><td> 7            </td><td>1068.04       </td><td>55728.80      </td></tr>\n",
       "\t<tr><td>CUS-3026014945</td><td>14            </td><td>2840.15       </td><td>74097.48      </td></tr>\n",
       "\t<tr><td>CUS-860700529 </td><td>14            </td><td>1808.62       </td><td>47185.60      </td></tr>\n",
       "</tbody>\n",
       "</table>\n"
      ],
      "text/latex": [
       "\\begin{tabular}{r|llll}\n",
       " customer\\_id & freq & level & annual\\_salary\\\\\n",
       "\\hline\n",
       "\t CUS-2487424745 &  7             & 1013.67        & 52891.85      \\\\\n",
       "\t CUS-2142601169 &  7             & 1002.13        & 52289.71      \\\\\n",
       "\t CUS-1614226872 &  7             &  892.09        & 46547.98      \\\\\n",
       "\t CUS-4123612273 &  7             & 1068.04        & 55728.80      \\\\\n",
       "\t CUS-3026014945 & 14             & 2840.15        & 74097.48      \\\\\n",
       "\t CUS-860700529  & 14             & 1808.62        & 47185.60      \\\\\n",
       "\\end{tabular}\n"
      ],
      "text/markdown": [
       "\n",
       "| customer_id | freq | level | annual_salary |\n",
       "|---|---|---|---|\n",
       "| CUS-2487424745 |  7             | 1013.67        | 52891.85       |\n",
       "| CUS-2142601169 |  7             | 1002.13        | 52289.71       |\n",
       "| CUS-1614226872 |  7             |  892.09        | 46547.98       |\n",
       "| CUS-4123612273 |  7             | 1068.04        | 55728.80       |\n",
       "| CUS-3026014945 | 14             | 2840.15        | 74097.48       |\n",
       "| CUS-860700529  | 14             | 1808.62        | 47185.60       |\n",
       "\n"
      ],
      "text/plain": [
       "  customer_id    freq level   annual_salary\n",
       "1 CUS-2487424745  7   1013.67 52891.85     \n",
       "2 CUS-2142601169  7   1002.13 52289.71     \n",
       "3 CUS-1614226872  7    892.09 46547.98     \n",
       "4 CUS-4123612273  7   1068.04 55728.80     \n",
       "5 CUS-3026014945 14   2840.15 74097.48     \n",
       "6 CUS-860700529  14   1808.62 47185.60     "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "head(df_inc)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA0gAAANICAMAAADKOT/pAAAAMFBMVEUAAABNTU1oaGh8fHyM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\ngEgAAiASgACIBCDAfzZw07SSsIKcAAAAAElFTkSuQmCC",
      "text/plain": [
       "Plot with title \"Histogram of customers' annual salary\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Visualise the distribution of customers' annual salary\n",
    "hist(df_inc$annual_salary[!is.na(df_inc$annual_salary)], breaks = c(seq(28000, 140000, by = 10000)), main = \"Histogram of customers' annual salary\", xlab = \"Income ($)\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Feature engineering"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Create a dataframe to store relevant features for customers\n",
    "df_cus = df_csmp %>% select(customer_id, gender, age, amount, date, balance) %>% group_by(customer_id) %>% mutate(avg_no_weekly_trans = round(7*n()/length(unique(df$date)), 0), max_amt = max(amount), no_large_trans = sum(amount > 100), use_no_day = length(unique(date)), avg_trans_amt = mean(amount, na.rm = TRUE), med_bal = median(balance, na.rm = TRUE)) %>% select(-c(\"amount\", \"date\", \"balance\")) %>% unique() \n",
    "\n",
    "# Create additional features\n",
    "df_cus$age_below20 = ifelse(df_cus$age < 20, 1, 0)\n",
    "df_cus$age_btw20n40 = ifelse(df_cus$age >= 20 & df_cus$age < 40, 1, 0)\n",
    "df_cus$age_btw40n60 = ifelse(df_cus$age >= 40 & df_cus$age < 60, 1, 0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table>\n",
       "<thead><tr><th scope=col>customer_id</th><th scope=col>State</th></tr></thead>\n",
       "<tbody>\n",
       "\t<tr><td>CUS-1005756958</td><td>QLD           </td></tr>\n",
       "\t<tr><td>CUS-1117979751</td><td>WA            </td></tr>\n",
       "\t<tr><td>CUS-1140341822</td><td>VIC           </td></tr>\n",
       "\t<tr><td>CUS-1147642491</td><td>NSW           </td></tr>\n",
       "\t<tr><td>CUS-1196156254</td><td>SA            </td></tr>\n",
       "\t<tr><td>CUS-1220154422</td><td>NSW           </td></tr>\n",
       "</tbody>\n",
       "</table>\n"
      ],
      "text/latex": [
       "\\begin{tabular}{r|ll}\n",
       " customer\\_id & State\\\\\n",
       "\\hline\n",
       "\t CUS-1005756958 & QLD           \\\\\n",
       "\t CUS-1117979751 & WA            \\\\\n",
       "\t CUS-1140341822 & VIC           \\\\\n",
       "\t CUS-1147642491 & NSW           \\\\\n",
       "\t CUS-1196156254 & SA            \\\\\n",
       "\t CUS-1220154422 & NSW           \\\\\n",
       "\\end{tabular}\n"
      ],
      "text/markdown": [
       "\n",
       "| customer_id | State |\n",
       "|---|---|\n",
       "| CUS-1005756958 | QLD            |\n",
       "| CUS-1117979751 | WA             |\n",
       "| CUS-1140341822 | VIC            |\n",
       "| CUS-1147642491 | NSW            |\n",
       "| CUS-1196156254 | SA             |\n",
       "| CUS-1220154422 | NSW            |\n",
       "\n"
      ],
      "text/plain": [
       "  customer_id    State\n",
       "1 CUS-1005756958 QLD  \n",
       "2 CUS-1117979751 WA   \n",
       "3 CUS-1140341822 VIC  \n",
       "4 CUS-1147642491 NSW  \n",
       "5 CUS-1196156254 SA   \n",
       "6 CUS-1220154422 NSW  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Investigate the state where customers live\n",
    "# Assume they live where most transactions occured\n",
    "df_region = df_csmp %>% group_by(customer_id, merchant_state) %>% summarise(trans_count = n()) %>% group_by(customer_id) %>% mutate(no_state = n()) %>% filter(trans_count == max(trans_count))\n",
    "\n",
    "# For equal number of transactions between multiple states, pick the most likely state\n",
    "n_occur = data.frame(table(df_region$customer_id))\n",
    "cus_id_rep = n_occur$Var1[n_occur$Freq > 1]\n",
    "\n",
    "state_by_cust_no = rev(names(sort(table(df_region$merchant_state), rev = TRUE)))\n",
    "t = data.frame(customer_id = cus_id_rep, merchant_state = NA)\n",
    "\n",
    "for (i in seq(length(cus_id_rep))){\n",
    "    s = df_region$merchant_state[df_region$customer_id == cus_id_rep[i]]\n",
    "    for (state in state_by_cust_no){\n",
    "        if (state %in% s){\n",
    "            t[i, 2] = state\n",
    "            break\n",
    "        }\n",
    "    }\n",
    "}\n",
    "\n",
    "df_region <- df_region[!(df_region$customer_id %in% cus_id_rep), c(1,2)] %>% as.data.frame() %>% rbind(t) %>% rename(State = merchant_state)\n",
    "\n",
    "head(df_region)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table>\n",
       "<thead><tr><th scope=col>customer_id</th><th scope=col>gender</th><th scope=col>age</th><th scope=col>avg_no_weekly_trans</th><th scope=col>max_amt</th><th scope=col>no_large_trans</th><th scope=col>use_no_day</th><th scope=col>avg_trans_amt</th><th scope=col>med_bal</th><th scope=col>age_below20</th><th scope=col>age_btw20n40</th><th scope=col>age_btw40n60</th><th scope=col>freq</th><th scope=col>level</th><th scope=col>annual_salary</th><th scope=col>State</th></tr></thead>\n",
       "<tbody>\n",
       "\t<tr><td>CUS-1005756958</td><td>F             </td><td>53            </td><td> 4            </td><td> 226.82       </td><td> 4            </td><td>25            </td><td>37.72625      </td><td> 1054.965     </td><td>0             </td><td>0             </td><td>1             </td><td> 7            </td><td> 970.47       </td><td> 50637.74     </td><td>QLD           </td></tr>\n",
       "\t<tr><td>CUS-1117979751</td><td>M             </td><td>21            </td><td> 4            </td><td>2885.85       </td><td> 1            </td><td>31            </td><td>76.45808      </td><td> 8444.820     </td><td>0             </td><td>1             </td><td>0             </td><td>14            </td><td>3578.65       </td><td> 93364.42     </td><td>WA            </td></tr>\n",
       "\t<tr><td>CUS-1140341822</td><td>M             </td><td>28            </td><td> 5            </td><td>1271.46       </td><td> 5            </td><td>33            </td><td>67.53138      </td><td> 6820.260     </td><td>0             </td><td>1             </td><td>0             </td><td>14            </td><td>1916.51       </td><td> 50000.38     </td><td>VIC           </td></tr>\n",
       "\t<tr><td>CUS-1147642491</td><td>F             </td><td>34            </td><td> 6            </td><td> 432.92       </td><td> 5            </td><td>39            </td><td>51.12829      </td><td> 9572.950     </td><td>0             </td><td>1             </td><td>0             </td><td> 7            </td><td>1711.39       </td><td> 89297.89     </td><td>NSW           </td></tr>\n",
       "\t<tr><td>CUS-1196156254</td><td>F             </td><td>34            </td><td>13            </td><td> 390.67       </td><td> 6            </td><td>65            </td><td>30.31049      </td><td>22055.110     </td><td>0             </td><td>1             </td><td>0             </td><td>14            </td><td>3903.73       </td><td>101845.53     </td><td>SA            </td></tr>\n",
       "\t<tr><td>CUS-1220154422</td><td>F             </td><td>25            </td><td> 4            </td><td> 168.74       </td><td>13            </td><td>34            </td><td>65.92896      </td><td> 8476.365     </td><td>0             </td><td>1             </td><td>0             </td><td>14            </td><td>2282.36       </td><td> 59545.14     </td><td>NSW           </td></tr>\n",
       "</tbody>\n",
       "</table>\n"
      ],
      "text/latex": [
       "\\begin{tabular}{r|llllllllllllllll}\n",
       " customer\\_id & gender & age & avg\\_no\\_weekly\\_trans & max\\_amt & no\\_large\\_trans & use\\_no\\_day & avg\\_trans\\_amt & med\\_bal & age\\_below20 & age\\_btw20n40 & age\\_btw40n60 & freq & level & annual\\_salary & State\\\\\n",
       "\\hline\n",
       "\t CUS-1005756958 & F              & 53             &  4             &  226.82        &  4             & 25             & 37.72625       &  1054.965      & 0              & 0              & 1              &  7             &  970.47        &  50637.74      & QLD           \\\\\n",
       "\t CUS-1117979751 & M              & 21             &  4             & 2885.85        &  1             & 31             & 76.45808       &  8444.820      & 0              & 1              & 0              & 14             & 3578.65        &  93364.42      & WA            \\\\\n",
       "\t CUS-1140341822 & M              & 28             &  5             & 1271.46        &  5             & 33             & 67.53138       &  6820.260      & 0              & 1              & 0              & 14             & 1916.51        &  50000.38      & VIC           \\\\\n",
       "\t CUS-1147642491 & F              & 34             &  6             &  432.92        &  5             & 39             & 51.12829       &  9572.950      & 0              & 1              & 0              &  7             & 1711.39        &  89297.89      & NSW           \\\\\n",
       "\t CUS-1196156254 & F              & 34             & 13             &  390.67        &  6             & 65             & 30.31049       & 22055.110      & 0              & 1              & 0              & 14             & 3903.73        & 101845.53      & SA            \\\\\n",
       "\t CUS-1220154422 & F              & 25             &  4             &  168.74        & 13             & 34             & 65.92896       &  8476.365      & 0              & 1              & 0              & 14             & 2282.36        &  59545.14      & NSW           \\\\\n",
       "\\end{tabular}\n"
      ],
      "text/markdown": [
       "\n",
       "| customer_id | gender | age | avg_no_weekly_trans | max_amt | no_large_trans | use_no_day | avg_trans_amt | med_bal | age_below20 | age_btw20n40 | age_btw40n60 | freq | level | annual_salary | State |\n",
       "|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n",
       "| CUS-1005756958 | F              | 53             |  4             |  226.82        |  4             | 25             | 37.72625       |  1054.965      | 0              | 0              | 1              |  7             |  970.47        |  50637.74      | QLD            |\n",
       "| CUS-1117979751 | M              | 21             |  4             | 2885.85        |  1             | 31             | 76.45808       |  8444.820      | 0              | 1              | 0              | 14             | 3578.65        |  93364.42      | WA             |\n",
       "| CUS-1140341822 | M              | 28             |  5             | 1271.46        |  5             | 33             | 67.53138       |  6820.260      | 0              | 1              | 0              | 14             | 1916.51        |  50000.38      | VIC            |\n",
       "| CUS-1147642491 | F              | 34             |  6             |  432.92        |  5             | 39             | 51.12829       |  9572.950      | 0              | 1              | 0              |  7             | 1711.39        |  89297.89      | NSW            |\n",
       "| CUS-1196156254 | F              | 34             | 13             |  390.67        |  6             | 65             | 30.31049       | 22055.110      | 0              | 1              | 0              | 14             | 3903.73        | 101845.53      | SA             |\n",
       "| CUS-1220154422 | F              | 25             |  4             |  168.74        | 13             | 34             | 65.92896       |  8476.365      | 0              | 1              | 0              | 14             | 2282.36        |  59545.14      | NSW            |\n",
       "\n"
      ],
      "text/plain": [
       "  customer_id    gender age avg_no_weekly_trans max_amt no_large_trans\n",
       "1 CUS-1005756958 F      53   4                   226.82  4            \n",
       "2 CUS-1117979751 M      21   4                  2885.85  1            \n",
       "3 CUS-1140341822 M      28   5                  1271.46  5            \n",
       "4 CUS-1147642491 F      34   6                   432.92  5            \n",
       "5 CUS-1196156254 F      34  13                   390.67  6            \n",
       "6 CUS-1220154422 F      25   4                   168.74 13            \n",
       "  use_no_day avg_trans_amt med_bal   age_below20 age_btw20n40 age_btw40n60 freq\n",
       "1 25         37.72625       1054.965 0           0            1             7  \n",
       "2 31         76.45808       8444.820 0           1            0            14  \n",
       "3 33         67.53138       6820.260 0           1            0            14  \n",
       "4 39         51.12829       9572.950 0           1            0             7  \n",
       "5 65         30.31049      22055.110 0           1            0            14  \n",
       "6 34         65.92896       8476.365 0           1            0            14  \n",
       "  level   annual_salary State\n",
       "1  970.47  50637.74     QLD  \n",
       "2 3578.65  93364.42     WA   \n",
       "3 1916.51  50000.38     VIC  \n",
       "4 1711.39  89297.89     NSW  \n",
       "5 3903.73 101845.53     SA   \n",
       "6 2282.36  59545.14     NSW  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Merge all the features into a single dataframe\n",
    "df_cus = df_cus %>% merge(df_inc) %>% merge(df_region)\n",
    "head(df_cus)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Build a simple regression model to predict the annual salary for each customer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "\n",
       "Call:\n",
       "lm(formula = annual_salary ~ . - customer_id - level - freq, \n",
       "    data = df_cus)\n",
       "\n",
       "Residuals:\n",
       "   Min     1Q Median     3Q    Max \n",
       "-36365 -17552  -4231  13235  68180 \n",
       "\n",
       "Coefficients:\n",
       "                      Estimate Std. Error t value Pr(>|t|)  \n",
       "(Intercept)         -5.461e+03  3.910e+04  -0.140   0.8893  \n",
       "genderM              4.668e+03  5.581e+03   0.837   0.4053  \n",
       "age                  7.133e+02  4.938e+02   1.444   0.1524  \n",
       "avg_no_weekly_trans -8.705e+02  9.622e+02  -0.905   0.3682  \n",
       "max_amt              2.013e+00  3.787e+00   0.532   0.5964  \n",
       "no_large_trans       5.850e+02  1.108e+03   0.528   0.5989  \n",
       "use_no_day           7.104e+01  3.292e+02   0.216   0.8297  \n",
       "avg_trans_amt       -1.905e+02  1.270e+02  -1.500   0.1375  \n",
       "med_bal              1.624e-01  8.231e-02   1.973   0.0518 .\n",
       "age_below20          6.522e+04  3.147e+04   2.072   0.0413 *\n",
       "age_btw20n40         5.701e+04  2.663e+04   2.141   0.0352 *\n",
       "age_btw40n60         4.615e+04  2.140e+04   2.157   0.0339 *\n",
       "StateNT              1.003e+04  1.999e+04   0.502   0.6173  \n",
       "StateQLD             4.012e+02  1.084e+04   0.037   0.9706  \n",
       "StateSA             -8.225e+03  1.181e+04  -0.696   0.4882  \n",
       "StateVIC            -3.717e+03  6.670e+03  -0.557   0.5788  \n",
       "StateWA              6.052e+03  8.864e+03   0.683   0.4967  \n",
       "---\n",
       "Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1\n",
       "\n",
       "Residual standard error: 25330 on 83 degrees of freedom\n",
       "Multiple R-squared:  0.1506,\tAdjusted R-squared:  -0.01314 \n",
       "F-statistic: 0.9198 on 16 and 83 DF,  p-value: 0.55\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fit1 = lm(annual_salary ~. - customer_id - level - freq, data = df_cus)\n",
    "summary(fit1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Start:  AIC=2043.34\n",
      "annual_salary ~ (customer_id + gender + age + avg_no_weekly_trans + \n",
      "    max_amt + no_large_trans + use_no_day + avg_trans_amt + med_bal + \n",
      "    age_below20 + age_btw20n40 + age_btw40n60 + freq + level + \n",
      "    State) - customer_id - level - freq\n",
      "\n",
      "                      Df  Sum of Sq        RSS    AIC\n",
      "- State                5 1409188476 5.4675e+10 2036.0\n",
      "- use_no_day           1   29889747 5.3296e+10 2041.4\n",
      "- no_large_trans       1  178928443 5.3445e+10 2041.7\n",
      "- max_amt              1  181364785 5.3447e+10 2041.7\n",
      "- gender               1  449087243 5.3715e+10 2042.2\n",
      "- avg_no_weekly_trans  1  525283699 5.3791e+10 2042.3\n",
      "<none>                              5.3266e+10 2043.3\n",
      "- age                  1 1338998926 5.4605e+10 2043.8\n",
      "- avg_trans_amt        1 1443399306 5.4709e+10 2044.0\n",
      "- med_bal              1 2498044638 5.5764e+10 2045.9\n",
      "- age_below20          1 2756070370 5.6022e+10 2046.4\n",
      "- age_btw20n40         1 2941567603 5.6208e+10 2046.7\n",
      "- age_btw40n60         1 2984866812 5.6251e+10 2046.8\n",
      "\n",
      "Step:  AIC=2035.95\n",
      "annual_salary ~ gender + age + avg_no_weekly_trans + max_amt + \n",
      "    no_large_trans + use_no_day + avg_trans_amt + med_bal + age_below20 + \n",
      "    age_btw20n40 + age_btw40n60\n",
      "\n",
      "                      Df  Sum of Sq        RSS    AIC\n",
      "- use_no_day           1   32525200 5.4708e+10 2034.0\n",
      "- no_large_trans       1  277837666 5.4953e+10 2034.5\n",
      "- max_amt              1  410237277 5.5085e+10 2034.7\n",
      "- avg_no_weekly_trans  1  557306627 5.5233e+10 2035.0\n",
      "- gender               1  657168851 5.5332e+10 2035.2\n",
      "<none>                              5.4675e+10 2036.0\n",
      "- age                  1 1411231861 5.6086e+10 2036.5\n",
      "- avg_trans_amt        1 1665356438 5.6341e+10 2037.0\n",
      "- med_bal              1 2403969266 5.7079e+10 2038.2\n",
      "- age_below20          1 3209856901 5.7885e+10 2039.7\n",
      "- age_btw20n40         1 3386009505 5.8061e+10 2040.0\n",
      "- age_btw40n60         1 3470570275 5.8146e+10 2040.1\n",
      "\n",
      "Step:  AIC=2034.01\n",
      "annual_salary ~ gender + age + avg_no_weekly_trans + max_amt + \n",
      "    no_large_trans + avg_trans_amt + med_bal + age_below20 + \n",
      "    age_btw20n40 + age_btw40n60\n",
      "\n",
      "                      Df  Sum of Sq        RSS    AIC\n",
      "- no_large_trans       1  441927188 5.5150e+10 2032.8\n",
      "- max_amt              1  534050281 5.5242e+10 2033.0\n",
      "- gender               1  679175406 5.5387e+10 2033.2\n",
      "- avg_no_weekly_trans  1  840090187 5.5548e+10 2033.5\n",
      "<none>                              5.4708e+10 2034.0\n",
      "- age                  1 1446825508 5.6155e+10 2034.6\n",
      "- avg_trans_amt        1 2118620255 5.6826e+10 2035.8\n",
      "- med_bal              1 2427738882 5.7135e+10 2036.3\n",
      "- age_below20          1 3314542362 5.8022e+10 2037.9\n",
      "- age_btw20n40         1 3451929846 5.8160e+10 2038.1\n",
      "- age_btw40n60         1 3488988167 5.8197e+10 2038.2\n",
      "\n",
      "Step:  AIC=2032.81\n",
      "annual_salary ~ gender + age + avg_no_weekly_trans + max_amt + \n",
      "    avg_trans_amt + med_bal + age_below20 + age_btw20n40 + age_btw40n60\n",
      "\n",
      "                      Df  Sum of Sq        RSS    AIC\n",
      "- max_amt              1  403823833 5.5554e+10 2031.5\n",
      "- avg_no_weekly_trans  1  492643567 5.5642e+10 2031.7\n",
      "- gender               1  606221363 5.5756e+10 2031.9\n",
      "<none>                              5.5150e+10 2032.8\n",
      "- age                  1 1467614903 5.6617e+10 2033.4\n",
      "- avg_trans_amt        1 1773642910 5.6923e+10 2034.0\n",
      "- med_bal              1 2388743489 5.7538e+10 2035.0\n",
      "- age_below20          1 3405713656 5.8555e+10 2036.8\n",
      "- age_btw20n40         1 3615668678 5.8765e+10 2037.2\n",
      "- age_btw40n60         1 3656154723 5.8806e+10 2037.2\n",
      "\n",
      "Step:  AIC=2031.54\n",
      "annual_salary ~ gender + age + avg_no_weekly_trans + avg_trans_amt + \n",
      "    med_bal + age_below20 + age_btw20n40 + age_btw40n60\n",
      "\n",
      "                      Df  Sum of Sq        RSS    AIC\n",
      "- avg_no_weekly_trans  1  251832600 5.5805e+10 2030.0\n",
      "- gender               1  565854646 5.6119e+10 2030.6\n",
      "<none>                              5.5554e+10 2031.5\n",
      "- avg_trans_amt        1 1453602653 5.7007e+10 2032.1\n",
      "- age                  1 1474353972 5.7028e+10 2032.2\n",
      "- med_bal              1 2177507763 5.7731e+10 2033.4\n",
      "- age_below20          1 3444980460 5.8998e+10 2035.6\n",
      "- age_btw40n60         1 3515861596 5.9069e+10 2035.7\n",
      "- age_btw20n40         1 3657421264 5.9211e+10 2035.9\n",
      "\n",
      "Step:  AIC=2030\n",
      "annual_salary ~ gender + age + avg_trans_amt + med_bal + age_below20 + \n",
      "    age_btw20n40 + age_btw40n60\n",
      "\n",
      "                Df  Sum of Sq        RSS    AIC\n",
      "- gender         1  684026341 5.6489e+10 2029.2\n",
      "<none>                        5.5805e+10 2030.0\n",
      "- avg_trans_amt  1 1255345047 5.7061e+10 2030.2\n",
      "- age            1 1446593647 5.7252e+10 2030.6\n",
      "- med_bal        1 2220625835 5.8026e+10 2031.9\n",
      "- age_below20    1 3354971048 5.9160e+10 2033.8\n",
      "- age_btw40n60   1 3399288962 5.9205e+10 2033.9\n",
      "- age_btw20n40   1 3502412935 5.9308e+10 2034.1\n",
      "\n",
      "Step:  AIC=2029.21\n",
      "annual_salary ~ age + avg_trans_amt + med_bal + age_below20 + \n",
      "    age_btw20n40 + age_btw40n60\n",
      "\n",
      "                Df  Sum of Sq        RSS    AIC\n",
      "<none>                        5.6489e+10 2029.2\n",
      "- avg_trans_amt  1 1181877136 5.7671e+10 2029.3\n",
      "- age            1 1407448645 5.7897e+10 2029.7\n",
      "- med_bal        1 2665079074 5.9154e+10 2031.8\n",
      "- age_below20    1 3185266094 5.9675e+10 2032.7\n",
      "- age_btw40n60   1 3201136986 5.9691e+10 2032.7\n",
      "- age_btw20n40   1 3420325219 5.9910e+10 2033.1\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "\n",
       "Call:\n",
       "lm(formula = annual_salary ~ age + avg_trans_amt + med_bal + \n",
       "    age_below20 + age_btw20n40 + age_btw40n60, data = df_cus)\n",
       "\n",
       "Coefficients:\n",
       "  (Intercept)            age  avg_trans_amt        med_bal    age_below20  \n",
       "   -8008.1390       713.6014      -114.2299         0.1625     67111.5662  \n",
       " age_btw20n40   age_btw40n60  \n",
       "   59001.1081     46207.0757  \n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "MASS::stepAIC(fit1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "\n",
       "Call:\n",
       "lm(formula = annual_salary ~ age + avg_trans_amt + med_bal + \n",
       "    age_below20 + age_btw20n40 + age_btw40n60, data = df_cus)\n",
       "\n",
       "Residuals:\n",
       "   Min     1Q Median     3Q    Max \n",
       "-36951 -17459  -3556  13641  69382 \n",
       "\n",
       "Coefficients:\n",
       "                Estimate Std. Error t value Pr(>|t|)  \n",
       "(Intercept)   -8.008e+03  3.555e+04  -0.225   0.8223  \n",
       "age            7.136e+02  4.688e+02   1.522   0.1313  \n",
       "avg_trans_amt -1.142e+02  8.189e+01  -1.395   0.1664  \n",
       "med_bal        1.625e-01  7.758e-02   2.095   0.0389 *\n",
       "age_below20    6.711e+04  2.931e+04   2.290   0.0243 *\n",
       "age_btw20n40   5.900e+04  2.486e+04   2.373   0.0197 *\n",
       "age_btw40n60   4.621e+04  2.013e+04   2.296   0.0239 *\n",
       "---\n",
       "Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1\n",
       "\n",
       "Residual standard error: 24650 on 93 degrees of freedom\n",
       "Multiple R-squared:  0.0992,\tAdjusted R-squared:  0.04109 \n",
       "F-statistic: 1.707 on 6 and 93 DF,  p-value: 0.128\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Backwards model selection together with stepAIC yield the most appropriate model\n",
    "fit2 = lm(formula = annual_salary ~ age + avg_trans_amt + med_bal + age_below20 + age_btw20n40 + age_btw40n60, data = df_cus)\n",
    "summary(fit2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA0gAAANICAMAAADKOT/pAAAAMFBMVEUAAABNTU1oaGh8fHyM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\nJEDgHwQJo+Ga0hYIAAAAAElFTkSuQmCC",
      "text/plain": [
       "plot without title"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Examine the residuals to capture any missed relationships \n",
    "plot(fit2$residuals, ylab = \"Residual\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "23767.4922294208"
      ],
      "text/latex": [
       "23767.4922294208"
      ],
      "text/markdown": [
       "23767.4922294208"
      ],
      "text/plain": [
       "[1] 23767.49"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "rmse(fit2$fitted.values, df_cus$annual_salary)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The RMSE of the model over the whole dataset is over $20,000 which indicates the inaccuracy of the model. The model's adjust R-squared also shows that it only explains about 4% of variation  in customers' annual salary. It is thus risky to use this linear model to predict customers' income bracket. More data is required to develop a more reliable model. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Build a decision-tree based model to predict salary"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA0gAAANICAMAAADKOT/pAAAAMFBMVEUAAABNTU1oaGh8fHyM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",
      "text/plain": [
       "Plot with title \"Regression tree for annual salary\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Split into training and test datasets\n",
    "smp_size = floor(0.75 * nrow(df_cus))\n",
    "\n",
    "# Set the seed to make partition reproducible\n",
    "set.seed(123)\n",
    "train_ind = sample(seq_len(nrow(df_cus)), size = smp_size)\n",
    "\n",
    "df_cus_train = df_cus[train_ind, ]\n",
    "df_cus_test = df_cus[-train_ind, ]\n",
    "\n",
    "fit3 = rpart(annual_salary ~ gender + age + avg_no_weekly_trans + max_amt + no_large_trans + use_no_day + avg_trans_amt + med_bal + age_below20 + age_btw20n40 + age_btw40n60 + State, method = \"anova\", data = df_cus_train)\n",
    "\n",
    "# Plot tree\n",
    "plot(fit3, uniform = TRUE, main = \"Regression tree for annual salary\")\n",
    "text(fit3, use.n = TRUE, all = TRUE, cex = 0.8)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "25809.4977586815"
      ],
      "text/latex": [
       "25809.4977586815"
      ],
      "text/markdown": [
       "25809.4977586815"
      ],
      "text/plain": [
       "[1] 25809.5"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Examine prediction accuracy\n",
    "rmse(actual = df_cus_test$annual_salary, predicted = predict(object = fit3, newdata = df_cus_test))"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "R",
   "language": "R",
   "name": "ir"
  },
  "language_info": {
   "codemirror_mode": "r",
   "file_extension": ".r",
   "mimetype": "text/x-r-source",
   "name": "R",
   "pygments_lexer": "r",
   "version": "3.6.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}