This chapter is about BKM Assigment.

1-Data Preparation

The Interbank Card Center (BKM) was established in 1990 with the partnership of 13 public and private Turkish banks for the purpose of providing solutions to the common problems and developing the rules and standards of credit and debit cards in Turkey, within the card payment system.

The dataset we used in this analysis from BKM includes number and volume of transactions made with credit card and debit card with respect to merchandise category group in Turkey.

# Creating a function to filter year and month on the URL

FuncYearMonth <- 
  function(year,month) {
    url <- paste('https://bkm.com.tr/en/secilen-aya-ait-sektorel-gelisim/?filter_year=',year,'&filter_month=',month,'&List=Lis'
                 ,sep=''
                 ,collapse = NULL)
  return(url)
  }

list_all <- c('201701','201702','201703','201704','201705','201706','201707','201708','201709','201710','201711','201712','201801','201802','201803','201804','201805','201806','201807','201808','201809','201810','201811','201812','201901','201902','201903','201904','201905','201906')

raw_df_full <- ''
for (i in 1:length(list_all)) {
v_year = substr(list_all[i],1,4)
v_month = substr(list_all[i],5,6)

    url<-FuncYearMonth(v_year,v_month)
    
    page <- read_html(url)
    
    raw_df <- 
      page %>%
      html_nodes("table") %>%.[(4)] %>% 
      html_table(page, fill = TRUE,header = FALSE) %>% 
      as.data.frame() %>%
      slice(3:max(nrow(.)))  
    
    # Replacing N/A values with 0 and adding year and month columns
    raw_df <- 
      raw_df %>% 
      mutate_if(is.numeric,funs(ifelse(is.na(.),0,.))) %>%  
      mutate(year = v_year,month = v_month)
    
    # Appending raw_df into raw_df_full in each iteration
    raw_df_full<-rbind(raw_df_full,raw_df)
}


# Adding column names
colnames(raw_df_full) <- c('merchant_category','cc_transaction_count','dc_transaction_count','cc_transaction_amount','dc_transaction_amount','year','month')

raw_df_full <- raw_df_full %>% slice(-c(1))

2- Analysis : Most DebitCard Spends of 3 Months of Every Year vs CreditCard Spends

This Analysis is about most debitcard spends 3 months of every year and compare same months creditcard spends.

# Most DebitCard Spends of 3 Months of Every Year
df_dbtcard <-
  raw_df_full %>% 
    select(year,month,dc_transaction_amount,cc_transaction_amount,dc_transaction_count,cc_transaction_count) %>%
    group_by(year,month) %>%
    summarise(debitcard_spend = sum(as.numeric((gsub(",","",dc_transaction_amount))),na.rm = T),
            creditcard_spend = sum(as.numeric((gsub(",","",cc_transaction_amount))),na.rm = T)
    ) %>%
    arrange(desc(debitcard_spend)) %>%
    mutate(rwn =row_number()) %>%
    filter(rwn<=3)

df_dbtcard
## # A tibble: 9 x 5
## # Groups:   year [3]
##   year  month debitcard_spend creditcard_spend   rwn
##   <chr> <chr>           <dbl>            <dbl> <int>
## 1 2019  05            810302.          314234.     1
## 2 2019  04            776785.          285674.     2
## 3 2018  09            767986.          394338.     1
## 4 2019  02            759635.          418690.     3
## 5 2018  11            661107.          456874.     2
## 6 2018  06            642760.          358358.     3
## 7 2017  12            600736.          594050.     1
## 8 2017  07            594307.          280250.     2
## 9 2017  10            582552.          504646.     3
df_dbtcard_pivot <- melt(df_dbtcard, id.vars = c("year" , "month"), 
                                    measure.vars = c("debitcard_spend", "creditcard_spend"))


ggplot(data = df_dbtcard_pivot, aes(x = paste(year,month) , y = value/1000 , group = variable)) + 
  #ggtitle("Plot of length \n by dose") + xlab("Dose (mg)") + ylab("Teeth length")
  geom_bar(aes(fill = paste(year,month)),stat = "identity") + scale_fill_hue() + theme(axis.text.x = element_text(angle = 30, hjust = 1))+
  #geom_bar(aes(fill = factor(..x.., labels = "paste(year,month)")), stat = "identity") + 
  labs(fill = "Year Month") + 
  facet_grid(~ variable) +
  scale_y_continuous("Spend(k)") +
  theme(axis.text.x = element_text(angle = 90, hjust = 1))

3- Analysis : Average Spent of Card Types

Yearly averages of the card types.

# Average Spent of Card Types
df_avg_spent <-
  raw_df_full %>% 
    select(year,dc_transaction_amount,cc_transaction_amount,dc_transaction_count,cc_transaction_count) %>%
    group_by(year) %>%
    summarise(debitcard_avg_spend = sum(as.numeric((gsub(",","",dc_transaction_amount))),na.rm=T)/sum(as.numeric((gsub(",","",dc_transaction_count))),na.rm = T),
     creditcard_avg_spend = sum(as.numeric((gsub(",","",cc_transaction_amount))),na.rm = T)/sum(as.numeric((gsub(",","",cc_transaction_count))),na.rm = T)
    )%>%
    arrange(desc(year)) %>%
    mutate(rwn =row_number()) %>%
    filter(rwn<=3)
df_avg_spent_pivot <- melt(df_avg_spent, id.vars = c("year"), 
                                    measure.vars = c("debitcard_avg_spend", "creditcard_avg_spend"))


ggplot(data = df_avg_spent_pivot, aes(x = year , y = value , group = variable)) + 
  #ggtitle("Plot of length \n by dose") + xlab("Dose (mg)") + ylab("Teeth length")
  geom_bar(aes(fill = year),stat = "identity") + scale_fill_hue() + theme(axis.text.x = element_text(angle = 30, hjust = 1))+
  #geom_bar(aes(fill = factor(..x.., labels = "year")), stat = "identity") + 
  labs(fill = "Year") + 
  facet_grid(~ variable) +
  scale_y_continuous("Spend") +
  theme(axis.text.x = element_text(angle = 90, hjust = 1))