My colleagues from the Arm Digital Signal Processing (DSP) software team have introduced a Python wrapper for the CMSIS-DSP library that is compatible with NumPy. The CMSIS-DSP library is a rich collection of DSP functions that Arm has optimized for the various Arm Cortex-M processors. CMSIS-DSP is widely used in the industry, and also enables optimized C code generation from various third-party tools.
Please read their blog here: https://developer.arm.com/architectures/instruction-sets/dsp-extensions/dsp-for-cortex-m/how-to-use-the-python-wrapper-for-csmsis-dsp-with-biquads-as-an-example
Hello Christopher,
I am haveing a similar problem (using jupyter notebook and typing in anaconda prompt):
- removed the folder from the OneDrive
- installed the numpy
- after going to the path and typing python setup.py build_ext –inplace -> I keep on having this errorAny suggestions, please?
Hello Tiago,
can you check if your ROOT variable is set to the right path? Take a look at the config.py file in the PythonWrapper folder and set the ROOT to the right path. It should be the path to the DSP folder. Currently there is ROOT = “..” and I this might not be robust enough.
Kind regards,
Christopher
Thank you for the fast reply.
That was really helpful.
Currently, I am having this error. Is it because I am using anaconda and/or should I move to visual studio?
Thank you,
here's a blog from Microsoft that talks about this for Anaconda users: https://devblogs.microsoft.com/python/unable-to-find-vcvarsall-bat/. I hope you will be able to download the package for Anaconda then.
Thanks again,
Last question (changes for pzthon2.7 with visual studio):
- the setup runs fine
- when opening a python notbook and running the import on the the same folder i get
ImportError: No module named cmsisdsp
Is it a problem with the setup.py?
Thanks in advance
Since it is an extension built in place, it is not installed in the Python folders. So if you are importing from the wrong working directory, the extension won’t be found. The working directory should be PythonWrapper
This is not ideal but we did not find the time for building and distributing a Python extension. So we are using the in place method.