r - Handling big data for apriori -

i'm trying run apriori arules package , running memory issues. far can read in data , split correct format doing basket analysis, when try convert data transaction type i'm getting error due size (about 800,000 rows).

is there packages can me handle size of data?

as((split(first2$product,f = first2$transaction_id)),"transactions") 

i've been stuck on week now, advice appreciated.

it depends on how many items (columns) have , how sparse data (items per transaction). apriori implementation in arules requires transactions reside in main memory. if can read in data as(data, "transactions") fails, can try directly create sparse ngcmatrix package matrix , use create itemmatrix (which transactions is).

see

r> class? ngcmatrix r> class? itemmatrix r> class? transactions 

unfortunately painful.