Reputation: 19
I am qualitatively coding a dataset based on theme. Each observation is allowed two themes, therefore I have two columns with the same variable list. When I run arules, it see "v1=alpha; v2=beta" as different item than "v1=beta;v2=alpha." As below,
| V1 | V2 |
| -------- | ----- |
| ALPHA | BETA |
| BETA | ALPHA |
Here's my code:
pr_itemset<-apriori(
pr_trans,parameter=list(
target="frequent",support=.001,minlen=2,maxlen=4))
Upvotes: 1
Views: 30
Reputation: 3075
These two rows are different. If you actually want the items to be ALPHA and BETA without the V1 and V2 because each row represents a set of items then you should start with a list of sets (represented as character vectors). The code would look like this:
library("arules")
mysets <- list(
c('ALPHA', 'BETA'),
c('BETA', 'ALPHA')
)
trans <- transactions(mysets)
inspect(trans)
items
[1] {ALPHA, BETA}
[2] {ALPHA, BETA}
identical(trans[1], trans[2])
[1] TRUE
Upvotes: 0