randn() function generates the same sequence whenever I called it.

Question

JangHo Cho 2015-8-4

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/232237-randn-function-generates-the-same-sequence-whenever-i-called-it

评论： JangHo Cho 2015-8-5

Hello,

I used MATLAB coder for converting .m files to C files.

All the things are O.K. except randn() function. The problem is that auto-generated randn() function generates the same sequence of numbers all the time. So I changed it by using 'srand(time(NULL))'; and 'rand/(double)RAND_MAX';

Are there any settings for MATLAB coder to generated a different version of randn() function?

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

Azzi Abdelmalek 2015-8-4

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/232237-randn-function-generates-the-same-sequence-whenever-i-called-it#answer_188138

Use rng function

for k=1:10
  rng(k)
  randn(1,10)
end

1 个评论
显示 -1更早的评论隐藏 -1更早的评论

JangHo Cho 2015-8-5

Thank you Azzi Abdelmalek. I understood by your code. it must be added to my function.

请先登录，再进行评论。

Answer 2

Mike Hosea 2015-8-4

1
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/232237-randn-function-generates-the-same-sequence-whenever-i-called-it#answer_188225

Every time you start MATLAB it will deliver the same sequence if you don't provide your own seed first. Your generated code must match this behavior. Consequently, if that isn't what you want, you must use the rng function to seed the generator. There is an example here on MATLAB Answers somewhere that illustrates using the C time() function to with rng to seed the generator.

1 个评论
显示 -1更早的评论隐藏 -1更早的评论

JangHo Cho 2015-8-5

Thank you for your kindness, Mike Hosea. I will search for the article you mentioned. Though I feel more comfortable to write directly in C code, I should search for it because in many cases we don't want to modify the auto-generated code lines. I thank both of you but there was only one chance to click the 'Accept this Answer' button. sorry. :)

请先登录，再进行评论。