The rules are sorted by lexicographical order.
Moreover, if there is a current data file, the rule generation procedure only keeps the rules with a maximum matching degree (on the whole data set) greater than a user given threshold. In this case the rules can optionally be sorted by decreasing order of cumulated weight.
Warning : the rule conclusions are initialized with a default value equal to one (crisp output), or to the first MF label (fuzzy output) . They can be reset by a separate procedure, described below.
The current FIS is not overwritten, a new FIS window pops up.