A framework for accessing data in models with many discrete states (abstract)
Mark C. Temte · 1990
In a class of applications, a model exists in which calculations depend upon a fixed set of independent real model parameters, say V1, V2, …, Vn. Each of the V's is a function of auxiliary data. The model often has a large number of discrete states, characterized by a set of discrete state parameters S1, S2, …, Sm. To calculate any particular V, a different function (algorithm and auxiliary data) may be needed for each state; typically, however, a function is shared among some group of states. It is desirable that only one copy of each function exist. The state may change frequently. Moreover, the state parameters are usually not independent; for example, choosing a particular value for one S may restrict permitted values for other S's. In an interactive application, inconsistent choices must be detected.