Matlab simulink error: Only finite double vector or matrix outputs are supported.

In my simulink model where i am using block "Interpreted Matlab function" and the function is:

function test = testFunc()

test = [int16(0) int16(0)]

i get error :

An error occurred while running the simulation and the simulation was terminated Caused by:

Error in 'test2/Interpreted MATLAB Function'. Evaluation of expression resulted in an invalid output. Only finite double vector or matrix outputs are supported

Edit: It crashes in second iteration

What am i doing wrong? Thanks for help

[budi raharjo]

I have tried by use coder.nullcopy() and it works.

function d=functionsx(s)
  a=[5 5 6; 7 6 8];
  b=a+s;
  d = coder.nullcopy(zeros(size(b)));
  d=b

Answer 1

[Birdman]

Why interpreted matlab function for a function like this? Can't you use instead Data Type Conversion block for it?

[Youda]

this is obviously my test function, real function should output two int16. after a long time i found out that it outputs this error when the two ints are 0. So i need to use interpreted matlab function

[Birdman]

Write your function in that block, but let the outputs be double. Then connect those outputs to Data Type Conversion block and convert them to int16.

[Youda]

ok, it works but why outputting two int16 zeros causes error?

[Birdman]

https://www.mathworks.com/help/simulink/slref/interpretedmatlabfunction.html

At the end of the page, you'll see Characteristics pane and there, Data Types->Double is written

[Youda]

thanks

Answer 2

[Xiaoxiong Zhang]

Maybe your function output is not a double type data, change your output to a double type data and see if it works? I have met the same problem before and I solve it out by changing my output data to double type using "output=double(output)". Hope this is helpful.

[Ting Liang]

Hi Xiaoxiong,

Excuse me.

I encountered the same problem now, I tried your methods, but it didn't work, could you please give some clues? Thanks. The problems are described as follows:

In simulink, I used the interpreted MATLAB function, to call for HEX_LT function, it works.

But if there is a subfunction (like SFHC) within the HEX_LT function, shown as follows:

function y = HEX_LT(para)

hex = SHFC (Pa_in, Ga, Qw, di, Tb);

Uo = 1/(1/ho + do^2 /(hex(i)* di^2));

... y = T;

end

I have tried several ways, but all didn't work, all told me that ' Evaluation of expression resulted in an invalid output. Only finite double vector or matrix outputs are supported', could you please help with this?

Thanks, and looking forward to your reply.

[Zihao Qi]

Hi，Ting Liang，

Excuse me，

I encountered the same problem too,have you solved this problem now?

[Omar Rateb Shehata]

good