Is double_t like C double? Could you get away with float instead?...
Here is an extract of the code :
...
If you go that route, check out the CMSIS DSP Software Library. It can make working in 16-bit or 32-bit fixed point numbers easier. It has arm_atan2_q31, and Filtering Functions.Without seeing to code it's hard to know where you can optimise it, but one suggestion might be to use fixed point calculation rather than floating point...
Statistics: Posted by carlk3 — Sat Feb 03, 2024 2:47 am