library(tidyverse)
library(completejourney)search()## [1] ".GlobalEnv" "package:completejourney"
## [3] "package:lubridate" "package:forcats"
## [5] "package:stringr" "package:dplyr"
## [7] "package:purrr" "package:readr"
## [9] "package:tidyr" "package:tibble"
## [11] "package:ggplot2" "package:tidyverse"
## [13] "package:stats" "package:graphics"
## [15] "package:grDevices" "package:utils"
## [17] "package:datasets" "package:methods"
## [19] "Autoloads" "package:base"ls("package:completejourney")## [1] "%<-%" "%>%" "campaign_descriptions"
## [4] "campaigns" "coupon_redemptions" "coupons"
## [7] "demographics" "get_data" "get_promotions"
## [10] "get_transactions" "products" "promotions_sample"
## [13] "transactions_sample"transactions <- get_transactions()
glimpse(transactions)## Rows: 1,469,307
## Columns: 11
## $ household_id <chr> "900", "900", "1228", "906", "906", "906", "906"…
## $ store_id <chr> "330", "330", "406", "319", "319", "319", "319",…
## $ basket_id <chr> "31198570044", "31198570047", "31198655051", "31…
## $ product_id <chr> "1095275", "9878513", "1041453", "1020156", "105…
## $ quantity <dbl> 1, 1, 1, 1, 2, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, …
## $ sales_value <dbl> 0.50, 0.99, 1.43, 1.50, 2.78, 5.49, 1.50, 1.88, …
## $ retail_disc <dbl> 0.00, 0.10, 0.15, 0.29, 0.80, 0.50, 0.29, 0.21, …
## $ coupon_disc <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ coupon_match_disc <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ week <int> 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, …
## $ transaction_timestamp <dttm> 2017-01-01 06:53:26, 2017-01-01 07:10:28, 2017-…dim(transactions)## [1] 1469307 11sum(is.na(transactions))## [1] 0colSums(is.na(transactions))## household_id store_id basket_id
## 0 0 0
## product_id quantity sales_value
## 0 0 0
## retail_disc coupon_disc coupon_match_disc
## 0 0 0
## week transaction_timestamp
## 0 0The transactions dataset contains 1,469,307 rows and 11 columns. It includes information about households, stores, products, quantities purchased, sales values, discounts, and the time of each transaction.
summary(transactions$sales_value)## Min. 1st Qu. Median Mean 3rd Qu. Max.
## 0.000 1.290 2.000 3.128 3.490 840.000The sales value is right-skewed. Most transactions have small sales amounts, but a few very large purchases increase the mean above the median.
summary(transactions$quantity)## Min. 1st Qu. Median Mean 3rd Qu. Max.
## 0.0 1.0 1.0 104.1 1.0 89638.0The quantity variable is strongly right-skewed. The median quantity purchased is 1 item, but the mean is much higher because a small number of transactions include very large quantities.
transactions_multi <- transactions %>%
filter(quantity > 1)
dim(transactions_multi)## [1] 302909 11Filtering the dataset to transactions with a quantity greater than 1 reduces the number of observations, showing that most transactions involve purchasing only one item.
transactions %>%
summarize(total_quantity = sum(quantity))## # A tibble: 1 × 1
## total_quantity
## <dbl>
## 1 152923108The total quantity purchased across all transactions is very large, reflecting the high volume of transactions in the dataset.
transactions %>%
group_by(household_id) %>%
summarize(total_quantity = sum(quantity)) %>%
head()## # A tibble: 6 × 2
## household_id total_quantity
## <chr> <dbl>
## 1 1 1131
## 2 10 7
## 3 100 122732
## 4 1000 30739
## 5 1001 1088
## 6 1002 250Grouping by household shows that total quantities purchased vary widely across households, indicating different shopping behaviors.
transactions_store309 <- transactions %>%
filter(store_id == "309", quantity > 1)
dim(transactions_store309)## [1] 1327 11Filtering to store 309 and transactions with quantity greater than one further reduces the dataset, showing that larger purchases at a single store are less common.
transactions %>%
arrange(desc(sales_value)) %>%
select(household_id, store_id, quantity, sales_value) %>%
head()## # A tibble: 6 × 4
## household_id store_id quantity sales_value
## <chr> <chr> <dbl> <dbl>
## 1 1609 412 3 840
## 2 346 415 5 632.
## 3 374 361 1 500.
## 4 1109 356 1 330.
## 5 2191 343 1 300.
## 6 1064 365 1 300.Sorting by sales value shows a small number of transactions with unusually large purchase amounts compared to the majority of transactions.
transactions %>%
group_by(store_id) %>%
summarize(total_sales = sum(sales_value)) %>%
arrange(desc(total_sales)) %>%
head()## # A tibble: 6 × 2
## store_id total_sales
## <chr> <dbl>
## 1 367 148170.
## 2 406 120904.
## 3 429 86884.
## 4 361 81196.
## 5 343 80701.
## 6 356 78874.Grouping by store reveals that total sales vary substantially across stores, with a small number of stores accounting for the highest total sales.
transactions %>%
distinct(household_id) %>%
nrow()## [1] 2469This shows how many unique households are represented in the transactions dataset. 2,469.
set.seed(123)
transactions_sample <- transactions %>%
sample_frac(0.10)
dim(transactions_sample)## [1] 146931 11A 10% sample was created to make the dataset more manageable while preserving its overall structure.
transactions %>%
slice(1:10)## # A tibble: 10 × 11
## household_id store_id basket_id product_id quantity sales_value retail_disc
## <chr> <chr> <chr> <chr> <dbl> <dbl> <dbl>
## 1 900 330 31198570044 1095275 1 0.5 0
## 2 900 330 31198570047 9878513 1 0.99 0.1
## 3 1228 406 31198655051 1041453 1 1.43 0.15
## 4 906 319 31198705046 1020156 1 1.5 0.29
## 5 906 319 31198705046 1053875 2 2.78 0.8
## 6 906 319 31198705046 1060312 1 5.49 0.5
## 7 906 319 31198705046 1075313 1 1.5 0.29
## 8 1058 381 31198676055 985893 1 1.88 0.21
## 9 1058 381 31198676055 988791 1 1.5 1.29
## 10 1058 381 31198676055 9297106 1 2.69 0
## # ℹ 4 more variables: coupon_disc <dbl>, coupon_match_disc <dbl>, week <int>,
## # transaction_timestamp <dttm>Slicing the data allows us to inspect specific rows to better understand the structure and values in the dataset.
transactions %>%
top_n(5, wt = sales_value) %>%
select(household_id, store_id, quantity, sales_value)## # A tibble: 6 × 4
## household_id store_id quantity sales_value
## <chr> <chr> <dbl> <dbl>
## 1 1609 412 3 840
## 2 374 361 1 500.
## 3 2191 343 1 300.
## 4 1109 356 1 330.
## 5 1064 365 1 300.
## 6 346 415 5 632.Selecting the top transactions by sales value highlights extreme purchase amounts compared to typical transactions.