How to force MATLAb to use 32bit floating-point numbers in a script?

Question

Mr M. 2018-4-17

1
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/395431-how-to-force-matlab-to-use-32bit-floating-point-numbers-in-a-script

评论： dpb 2018-4-17

Is it possible to use 32bit precision thru a script or part of the script without editing the whole code?

2 个评论
显示无隐藏无

John D'Errico 2018-4-17

No. You cannot simply tell MATLAB to change precision from outside the code.

dpb 2018-4-17

On top of which, you can't define a default precision for a script within a script or function file; you can only cast to a given precision on assignment and then ensure all variables in all related calculations are also of same class or follow other strict rules to avoid results being default doubles.

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Walter Roberson 2018-4-17

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/395431-how-to-force-matlab-to-use-32bit-floating-point-numbers-in-a-script#answer_315597

编辑：dpb 2018-4-17

Yes, but only on MS Windows

https://www.mathworks.com/matlabcentral/answers/47142-how-to-set-matlab-precision-to-single

1 个评论
显示 -1更早的评论隐藏 -1更早的评论

dpb 2018-4-17

Doesn't it only control the rounding mode? I couldn't demonstrate to my satisfaction that it actually did anything discernible in R2016b here; seems like in times gone by one could see the effect but I've forgotten the details of how/what did those many years ago...

I'm presuming the OP wants to actually have everything only take up single precision in memory as well???

请先登录，再进行评论。