Is there a way to specify all 'yes/1' responses in Apriori on RHS? (R Studio)

Question

Good afternoon,

I'm conducting Market Basket analysis at the moment and I have reduced the LHS rules down to the top 15 most frequently bought items, and this has worked well using the following code:

rules2 <- apriori(bb_pivot,
             parameter = list(minlen=2,maxlen=3,conf=0.95),
             appearance = list(lhs=c("Coffee=1",
                                     "Bread=1",
                                     "Tea=1",
                                     "Cake=1",
                                     "Pastry=1",
                                     "Sandwich=1",
                                     "Medialuna=1",
                                     "Hot chocolate=1",
                                     "Cookies=1",
                                     "Brownie=1",
                                     "Farm House=1",
                                     "Muffin=1",
                                     "Alfajores=1",
                                     "Juice=1",
                                     "Soup=1"),
                               default="rhs"))

However, I am only interested in the rules that are generated when other items are bought with these top 15 items, not the times when they were bought in the absence of another item. The rules I've generated with the above code mostly follow the below output:

        lhs         rhs                               support confidence  coverage    lift   count
[1]   {Cake=1} => {Medialuna=0}                     0.1001585 0.9643947  0.1038563 1.0279275  948 
[2]   {Cake=1} => {Farm House=0}                    0.1023772 0.9857579  0.1038563 1.0259730  969 
[3]   {Cake=1} => {Toast=0}                         0.1016376 0.9786368  0.1038563 1.0126596  962 
[4]   {Cake=1} => {Scandinavian=0}                  0.1023772 0.9857579  0.1038563 1.0152555  969 
[5]   {Cake=1} => {Truffles=0}                      0.1012150 0.9745677  0.1038563 0.9947463  958 

Is there a way for me to only select rules based on the RHS being 'yes/1', without coding for the 94 different items in my set on the RHS?

Answer 1

You have two options:

1. Template: You can add rhs = "yes/1" to the appearance list in your apriori call and remove default = "rhs" (see ? template).
2. Filtering: You can filter the rules you have by using rules3 <- subset(rules2, rhs %in% "yes/1") (see ? subset in package arules).