The new HackSpace Magazine Issue 30 presents using the new ulab (pronounced microlab) library for CircuitPython. ulab is a subset of the Python monster data crunching library numpy. An Adafruit CLUE board is used to demonstrate the power of fast mathematics on microcontrollers.
Python is fine for many uses, but when it comes to number crunching, it struggles. For regular Python,
designed to run on full computers, the solution comes in the form of NumPy and SciPy. These libraries contain number-crunching code that runs much quicker than is usually possible. This is because Python can work with code written in other, much faster, languages.While NumPy and SciPy don’t work in MicroPython and CircuitPython, there is a similar library that does – ulab. We’ll look at the CircuitPython version here.
It’s only available on the very latest release – 5.1 rc0 – or higher, so you’ll need to upgrade to this. Also,
it’s only compatible with Cortex-M4 ARM chips at the moment, which mostly means those boards based on SAMD51 or nRF52 series microcontrollers. You can download the latest firmware for any CircuitPython compatible device from circuitpython.org. We’ve used the Adafruit CLUE for this code as it has both a built-in microphone and screen, but you can use the same technique with any data that varies over time.
Read more, download PDF, buy now, subscribe.
Read the Adafruit Learning System guide on ulab
See Adafruit’s CLUE on Adafruit.com and at Digi-key.com.