From g, it retrieves an explanation that consists of (i) a factual decision rule, that corresponds to the path on the decision tree followed by the instance x to reach the decision y, and (ii) a set of counterfactual rules, which have a different classification w.r.t. y.