Yes, because it tests on a user-by-user basis. There's not the same notion of test/training set. Each user is split individually one at a time, according to the "at" parameter.
On Wed, May 30, 2012 at 10:16 AM, Daniel Quach <[email protected]> wrote: > I want to use the GenericRecommenderIRStatsEvaluator to get > precision/recall scores for my recommenders. > > I know with the other recommender evaluators, you can specify the split of > training data and test data. However, I see no such parameter for the IR > Stats evaluator. Am I missing something here, or is it not sensible to > split the data for this kind of evaluation?
