How to change the real_T in generated C code to be float type?

Question

Yunlong Wang 2016-3-27

1
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/275650-how-to-change-the-real_t-in-generated-c-code-to-be-float-type

评论： Rami Al Halabi 2016-3-27

Hi, I am now using Matlab2015b. I use the Code Generation of simulink to convert the algorithm inside Matlab Function into C code. In the generated code, I found that the real_T is float, as shown below

typedef doulbe real_T;

However, I hope that real_T could be float, so that the calculation time needed inside my ARM chip could be saved a lot.

I know that one method is specifying every variable with single() function, but there are so many variables inside my algorithm. So is there some other better methods? Thank you.

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Rami Al Halabi 2016-3-27

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/275650-how-to-change-the-real_t-in-generated-c-code-to-be-float-type#answer_215203

在 MATLAB Online 中打开

have you tried replacing double by float in that typedef statement?

typedef float real_T;

2 个评论
显示无隐藏无

Yunlong Wang 2016-3-27

编辑：Yunlong Wang 2016-3-27

Hi Rami, thank you. I have done that. I works well, but I have to modify the generated code each time after code generation. Is there some other better methods?

Rami Al Halabi 2016-3-27

No problem. Not sure if there is a setting in matlab that allows you to specify whether you want single or double precision. If you find any, please share it here :)

请先登录，再进行评论。