在R中找到重复的图案

Question

In R, I have a list of lists with data such as

1 "a" "b" "c"
2 "a" "a" "b" "d"
3 "a" "a"

I need to identify the commonly occurring patterns in the lists. For example, here, "a""a", and "a""b" are common. I tried using eclat(), but that doesn't allow for repeated values in each list. I then tried removing the duplicate values, but then I lose information (like "a""a" is a frequent pattern).

I also tried renaming the duplicate occurrences, but then "a""b" and "a""a""b" won't return "a""b" as a pattern, since the second list would be renamed to something like "a""a2""b".

Is there any better way to do this?

Update:

The strings in each list can be single characters or a string of characters. For example

1 "a+12" "bfd" "c"
2 "a+12" "a+12" "bfd" "d"
3 "a+12" "a+12" "a"

Here, "a+12" "bfd" and "a+12" "a+12 should be recognized as patterns

Answer 1

At least with the sample data, something like the following looks helpful:

#data in a 'list'
myls = list(x1 = c("a+12", "bfd", "c"), 
            x2 = c("a+12", "a+12", "bfd", "d"),
            x3 = c("a+12", "a+12", "a"))

pats = table(unlist(lapply(myls, 
           function(x) combn(seq_along(x), 2, 
                             function(i) paste(x[i[1]:i[2]], collapse = ";")))))
strsplit(names(pats[pats == max(pats)]), ";")
#[[1]]
#[1] "a+12" "a+12"
#
#[[2]]
#[1] "a+12" "bfd"