Search code examples
rarules

Transaction items are unintentionally shown as factor IDs using arules write() function

I have a list of transactions in a .csv file. For each row, the first column contains a "TransactionID". In the second columns contains the "ItemName" of a single item in that transaction. The third column contains the corresponding "ItemNumber".

I import those transactions with read.csv. 

trans = read.csv("MyFile.csv", header = TRUE, sep = ";")

Then I take the relevant columns and convert them into a list with factors.

TransList <- split(trans[,"ItemNames"], trans[,"TransactionID"])

Finally I coerce this list to transactions.

TransMat <- as(TransList, "transactions")

Now I want to write the contents of each "basket" into a file.

write(TransMat, file = "Basket.csv", sep = ", ", quote = FALSE)

Unfortunately, instead of writing the original "ItemNames" into the file, the FactorIDs are written into the file.

What really confuses me, is the fact that this does not happen when I use the "ItemNumbers" instead of the "ItemNames".

How can I write the original "ItemNames" into a .csv file instead of the FactorIDs?

Thanks!

Solution

  • I was able to solve my own question.

    Simply coercing to a data.frame before splitting did the trick.

    trans <-as.data.frame(trans)

    But honestly, I don't understand why this works... If anyone were to enlighten me in a comment I would greatly appreciate it.