Search code examples
rdata-miningarules

subset a-rules in R by length of lhs

I'm using the arules package in R to extract some association rules and want to filter by the length of the left hand side (lhs).

I tried using subset but without success.

This is my mode and attempt:

  trans<-read.transactions(file='file.csv', format="single", sep=",",cols=c("userid","target"))

  inspect(head(trans,3))

  rules<-apriori(trans,parameter=list(support=0.06, confidence=0.5,minlen=3,maxlen=6)) 

  # sort by support
  top.support <- sort(rules, decreasing = TRUE, na.last = NA, by = "support")

This is what I'm doing:

subset(rules,subset=length(lhs)==5) # not working. it return set of 0 rules

Any suggestions?

Solution

  • length gives you the number of rules. You need to use size instead.

    subset(rules,subset = size(lhs) == 5)