I am currently working on a discrete version of a S-function in order to make the execution process quicker (fixed step solver instead of ode1 or so).
For this I need to eliminate all derivatives in the function. Normally this is possible by allocating a static variable and reading out the current step size to determine the value change per step size.
But in this case I need to do this for an array of values and additionally the size is not fixed but determined by a parameter. So it's not possible to declare the array before the #define statements, because there is the parameter unknown yet.
I know it's possible to define Inputs, Outputs and Parameters with so far not known array size, so could I use sth like Rwork for this issue? And if yes, how does it work then?
solved it myself already. An easy way is to declare a pointer with "static double *Pointername" at the S-function Top before all the #defines.
Then the size can be determined by a parameter in the Initialize function with: "Pointername = malloc (Parametername * sizeof (double));"
And with that, S-functions can calculate arrays of derivatives without using the derivative function -> Fixed step Discrete Solver (which is quicker than ode1)can be used.