Abstract

A standard program starts its execution in an initial state, and terminates (if it ever does) in one of a set of final states. Its behaviour can be modelled by a binary relation between the initial and final states. The difference between the standard and the probabilistic semantics is that the former tells us which final states are or are not possible, whereas the latter tells us the probability with which they may occur. This paper presents a link between the probabilistic and the imperative programming using the weakest completion. We demonstrate how the probabilistic semantics can be derived directly from the standard relational one using the type embedding and healthiness condition of real programs. Carroll Morgan is an adjunct professor in the department of computer science at New South Wales University in Australia. He conducts research in the area of refinement theories and formal methods applied to software engineering and applications to parallel and distributed computing, ...

Citations