Custom processing block example (Python)

This is an example of a custom processing block, which you can load in the Edge Impulse studio. See the docs: Building custom processing blocks.

Additionally, this repo demonstrates how to use Octave inside Edge Impulse Studio. The key steps are

1. Install octave into the container (apt install octave liboctave-dev)

2. Install Octave package dependencies. Installing via the Dockerfile will save time and avoid having to wait for an install each time you train.

• Note, on Linux or inside containers, if a pacakge added to setup.m fails to build, you can also try to install it via apt install, example: apt install octave-control

See the following line in the Dockerfile

RUN octave setup.m

3. Invoke Octave from Python via the Oct2Py package

• Start your octave instance, and run your octave script file so your functions will be added to the Octave namespace. Do this outside of the generate_features function, so that your Octave instance will persist across samples

oc = oct2py.Oct2Py()

• Invoke functions like so: result = oc.my_func([1,2,3])
• Numpy arrays are also valid input