Maximum size of output not bounded in matlab function block for linspace error

I am using matlab function block in simulink and need to use linspace for use in my function. When I run the model, I get an error saying "Computed maximum size of the output of function 'linspace' is not bounded. Static memory allocation requires all sizes to be bounded. The computed size is [1 x :?]." How do I get around this? The code where I use linspace is below where 'd't is a single value calculated earlier in the model and is imported into the block and 'n' is a constant coming from a constants block.

vec = linspace(0,dt,n)

Answer 1

[Honglei Chen]

You said the value n is from a constant block, so do you ever need to change it? If the answer is no, then you can consider hard code it in the function. If you want to be able to change it from time to time but never need to change it during the model running, then you can also consider set it up as a nontunable dialog parameter so the engine knows that your vec has a fixed size.

HTH

[Andrew Poissant]

Sometimes I need to change the amount of points in linspace but I could hardcode it because I would only need to change it before the sim starts. The issue I believe is the dt variable. It is solved for earlier in the model using fsolve in an extrinsic matlab block so depending on the model conditions it could change therebye changing the range of vec.

[Honglei Chen]

Have you tried it? Based on your description, the output size is determined by N, not dt, so I don't think dt is the culprit for this error.

[Andrew Poissant]

You are right! My bad. But why would Simulink give me an error using a constants block? I preset the block to a value of 8 and do not change it so shouldn't Simulink know the size?

[Honglei Chen]

All MATLAB Function block knows is that this value is coming from the input, it has no idea whether the previous block is a constant or not so it has to prepare for the worst case. (Besides, even for a constant block, I believe you can still tune the parameter during the simulation)