R关联规则apriori - 抑制项目的分组

Question

I'm using the apriori function from the arules package to perform analyses for item association. After coercing the results into a datframe, I notice the output can be grouped like so in some instances:

 LHS                 RHS        Support        Confidence      Lift        Count
{Item1, Item2}      {Item3}      .84             0.99          12.3          6

Is there a way for force the function to perform the analysis just between two items in each transaction and yield the corresponding support, confidence, etc? In other words, can I force apriori to split the above result into something like:

 LHS           RHS        Support        Confidence      Lift       Count
{Item1}      {Item3}        a                b            c           d
{Item2}      {Item3}        e                f            g           h

Answer 1

maxlen in apriori does not act as a filter after mining. apriori creates frequent itemsets starting with length 1 then 2, 3, ... and maxlen stops the mining process.

minlen on the other hand is a filter because all shorter frequent itemsets have to be found first.

If you are asking if you can determine support and confidence of

{Item1} -> {Item3}
{Item2} -> {Item3}

by just using the support and confidence of

{Item1, Item2} -> {Item3}

then the answer is no. However, from the apriori property we know at least the following:

supp({Item1} -> {Item3}) >= supp({Item1, Item2} -> {Item3})
supp({Item2} -> {Item3}) >= supp({Item1, Item2} -> {Item3})