The fundamental difference between deductive and inductive argumentation is that deductive arguments yield certain conclusions, given their premises, but the premise/premises of an inductive argument only yields probability. In other words, if the premises of a deductive argument are true, then the conclusion is true. However, note that irksome word, if. So the conclusion of a deductive argument is only irrefragable when the premises are (necessarily) true. Given p, q follows; I have p, therefore I have q.
Another way that some explain inductive argumentation is by making a contrast between reasoning from general premises to specific conclusions (as in the case of deductive arguments) versus reasoning from concrete particulars to general conclusions (i.e., inductive arguments). I will not deal with the inadequacies of characterizing matters this way, but I merely want my readers to know the difference between deductive and inductive arguments for the purpose of grasping a problem that I have with Calvinism in toto.
Calvinism--at least, some Calvinists--reasons inductively from concrete particulars to general conclusions. For instance, let us assume that event1 (E1) represents an occurrence of evil that has a good outcome (i.e., God brings something good from the evil occurrence). Calvinist seem to reason that if God brings good from E1, E2, E3--then he also brings good out of En. However, I'm not sure that the reasoning holds up; after all, inductive arguments result in probable conclusions. Maybe Calvinists object that their reasons for believing that God brings good--or is able to bring good--from evil depends on more than rational arguments that are inductive. It is possible that the Calvinist is bypassing logic/reason and basis his/her argument on Scripture.
The foregoing reasoning notwithstanding, my comments are directed at the Calvinist, who uses actual events like the Holocaust or slavery to reason that if God brought good from evil in some instances, then one can infer that God brings good from all evil occurrences. My contention is that the logical entailment likely does not follow since one can't derive certainty from arguments that only yield probability.
David Duncombe spells out a corresponding line of reasoning here: https://daveduncombe.wordpress.com/2015/09/04/gods-glory-and-the-problem-of-induction/