Confidence implementation is not correct

[ferranmunoz]

Do you agree the following comment from the blog?
The otherItemsFrequency can be obtained from fList for the rules with a single item in the precondition.

But it seems to me that your confidence implementation is not corrrect still.
In yours code:
double confidence = (double)occurrence / firstFrequencyItem;

Let’s look at formula: conf(X => Y)=support(X&Y)/support(X)
In code, occurence is value that’s coresponds to X&Y set, but firstFrequencyItem is value that corresponds to Y. So you calculate support(X&Y)/support(Y). As you can see it’s not a confidence for rule X=>Y.

Implementation should look like
double confidence = (double)occurrence / otherItemsFrequency;

But how this otherItemsFrequency can be found in mahout output files?
My prectice shows that such frequencies can be found in frequentpatterns output but not for all rules.