casting without rounding using coder

4 次查看（过去 30 天）
Marc Barnada
Marc Barnada 2022-7-12
评论： Bjorn Gustavsson 2022-7-18
How can I manually trigger casting from a float/double to integer without involving any rounding while using coder?
LUT = single(rand(1,2000)*200);
for i=1:2000
ind = int32(LUT(i));
fprintf("%i\n",ind);
end
When the C++ code is generated it will be similar to:
int ind = static_cast<int>(rt_roundf_snf(LUT[i]));
I am aware that under certain circumstances coder is able to automatically figure out the optimal datatypes for indexing inside for loops resulting then in the desired result:
int ind = static_cast<int>(LUT[i]);

请先登录，再回答此问题。

回答（1 个）

Bjorn Gustavsson
Bjorn Gustavsson 2022-7-12
Well, the only interpretation I can thing of that matches the "casting without rounding" in a sensible way is:
ind = int32(floor(LUT2(i));
That way ind will be the integer part of whatever LUT2 is. But this is still "rounding-off the fractional part".
HTH
  4 个评论
显示 2更早的评论
Marc Barnada
Marc Barnada 2022-7-18
I want to have highly efficient C++ code, and that extra call to rt_roundf_snf has a big impact in the timings of a Look-Up-Table (LUT) function.
Bjorn Gustavsson
Bjorn Gustavsson 2022-7-18
If it is a limited point-of-use and a one-time effort would it be possible to manually edit the generated code?

请先登录，再进行评论。

类别

在 Help Center 和 File Exchange 中查找有关 MATLAB Coder 的更多信息

标签

产品


版本

R2021a

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by