I've been working on rewriting my code that worked with data.frames to work with ffdf. I had two columns, and after a lot of fuss I've managed to do a split and get a list with the following look:
data=
$A
1 2 3
$B
4 5 6
where A,B are the "baskets" or groupings, and "1 2 3" specific grouped items. What I want now is to convert these to transactions and hopefully manage to do an apriori. I've tried the simple
as(i, "transaction")
which worked well when "data" was generated from a data.frame but now it produces an error:
Error in as(data, "transactions") :
no method or default for coercing “list” to “transactions”
I've seen that duplicate items can cause these problems, so I've eliminated those but the error remains.
Should be no problem:
library(arules)
data <- list(A=1:3, B=4:6)
showMethods("coerce", classes="transactions")
# Function: coerce (package methods)
# from="data.frame", to="transactions"
# from="list", to="transactions"
# from="matrix", to="transactions"
# from="ngCMatrix", to="transactions"
# from="tidLists", to="transactions"
# from="transactions", to="data.frame"
# from="transactions", to="list"
# from="transactions", to="matrix"
# from="transactions", to="tidLists"
class(data)
# [1] "list"
as(data, "transactions")
# transactions in sparse format with
# 2 transactions (rows) and
# 6 items (columns)
Also note that you wrote as(i, "transaction") and not as(i, "transactions").