Variability and Subtyping

The variability of an expression governs the degree of freedom with which the expression may vary during simulation. Variability is a specific property but has clear connections with the notion of data type. The example classes below illustrates the subtyping and variability rules for declaration equations and normal equations.

1 Constants

This is not allowed, since p1 + 2 is not a constant expression

model Constants parameter Real p1 = 1; constant Real c1 = p1 + 2; // Not allowed - p1 has constant variability end Constants;

2 Simulation of Constants

Below the error message is shown, after trying to simulate the Constants model.

simulate( Constants ) //?? Error - should have emitted variability error

3 ConstantsCorrect

This is correct, since p1 + 2 has the same variability as p1.

model ConstantsCorrect parameter Real p1 = 1; parameter Real p2 = p1 + 2; // Fine! - p1 + 2 has same variability end ConstantsCorrect;

4 Simulation of ConstantsCorrect

simulate( ConstantsCorrect )

5 TestEquations

The equation in TestEquations is not correct, since b is a discrete time and noEvent(x > 1) is a continuous time expression.

model TestEquations Constants c1(p1 = 3); Constants c2(p2 = 7); // Fine, declaration equation can be modified. Boolean b; Real x; equation b = noEvent(x > 1); end TestEquations;

6 Simulation of TestEquations

Below the error message is shown, after trying to simulate the TestEquations model.

simulate( TestEquations )