Aggregating Categorical Variable Values in a single variable in R

Question

I have a dataset with the following variables - Bill_Number,Item_Name,Quantity. A single Bill_Number can have multiple Item_Names under it. I am trying to summarize it by Bill_Number and Quantity wherein the Item_Names are concatenated into a single variable for each bill number. The solutions here :Combine several row variables work well for a limited and pre-defined frame (like given below), but I have hundreds of Item_Names. I am sure there is a simpler way to do this. Can anyone help please?

Data (sample only):

BillN<-c('B1','B1','B1','B1','B2','B2','B2','B2','B3','B3','B3','B3')

Item_Name<-c('Prod A','Prod B','Prod C','Prod D','Prod A','Prod B','Prod C','Prod D','Prod A','Prod B','Prod C','Prod D') # going on to Product(n)

Quantity<-c(1,2,1,2,1,2,1,1,1,2,1,1)

Output: using ply::ddply

ddply(within(Dummy1, {
  Item_Name <- ifelse(Item_Name %in% c('Prod A','Prod B','Prod C'), 'Prod A + Prod B + Prod C', 'Prod D')
}), .(BillN, Item_Name), summarise, Count=sum(Quantity))

This depends on 'Prod D' being a separate row, whereas I need a combination of all the products for a given number and the overall quantity for that bill number.

The output I require is in this format:

BillN | Item_Name | Quantity

B1 | Prod A + Prod B + Prod C + Prod D | 4

B2 | Prod A + Prod B + Prod C + Prod D | 4

B3 | Prod A + Prod B + Prod C + Prod D | 4

Whereas the output I get from the code above is:

BillN | Item_Name | Quantity

B1 | Prod A + Prod B + Prod C | 4

B1 | Prod D | 2

B2 | Prod A + Prod B + Prod C | 4

B2 | Prod D | 2

B3 | Prod A + Prod B + Prod C | 4

B3 | Prod D | 2

Extending the scenario, what happens if I have more variables such as the Brand, Category, Sub_category for each Item_Name and the Sales_Amount as a numeric variable to further sum it by?

Please let me know if you need more information or have questions.

Thanks, Raoul

Answer 1

Here is a quick solution using data.table package:

Step1: Create the data.table

library(data.table)

DT <- data.table(
  BillN=c('B1','B1','B1','B1','B2','B2','B2','B2','B3','B3','B3','B3'),
  Item_Name=c('Prod A','Prod B','Prod C','Prod D','Prod A','Prod B','Prod C','Prod D','Prod A','Prod B','Prod C','Prod D'), # going on to Product(n)
  Quantity=c(1,2,1,2,1,2,1,1,1,2,1,1)
)

Step2: Set appropriate key:

setkey(DT,BillN)

Step3: Make sure that the string vector Item_Name is not a factor.

DT[,Item_Name := as.character(Item_Name)]

Step4: Perform the operation by key

DT[,list(Item_Name =paste(Item_Name,collapse=" + "),
         Quantity=sum(Quantity)),
   by=key(DT)]

You can use the result as it is, or collapse each row! Which is simple!!

Answer 2

Here's an example summing both the Quantity variables by BillN and the count of BillNs as your expected output wants:

Dummy1 <- data.frame(BillN, Item_Name, Quantity)

ldply(by(Dummy1, Dummy1$BillN, simplify=TRUE, FUN=function(x) {

  BillN <- x[1,]$BillN
  Quantity <- sum(x$Quantity)
  Item_Name <- paste(as.character(unique(x$Item_Name)), sep="", collapse=" + ")
  BillCt <- nrow(x)

  return(data.frame(BillN, Item_Name, Quantity, BillCt))

}))[2:5]

##  BillN                         Item_Name Quantity BillCt
## 1    B1 Prod A + Prod B + Prod C + Prod D        6      4
## 2    B2 Prod A + Prod B + Prod C + Prod D        5      4
## 3    B3 Prod A + Prod B + Prod C + Prod D        5      4