1

I am trying to properly perform FFT and calculate frequencies on my PICO using CircuitPython's (CP) ULAB library: ULAB FFT

With traditional Python using NUMPY and SCIPY, this is easily achievable, however it seems there is not a straightforward way to to calculate the frequencies of the FFT values in CP. There is no specific function for this under the ULAB FFT section; it also mentions ulab.extras.spectrum, but I cannot seem to find that function in the module attributes nor does it import. I figured maybe the magnitudes will be sufficient or of some better use.

Here is an example code

import ulab
import ulab.fft
   
spectrogram1 = ulab.fft.spectrogram(raw_adc_data_array)
spectrogram1 = ulab.vector.log(spectrogram1 + 1e-7)
spectrogram1 = spectrogram1[1:(samples_per_second//2)-1]

I have an ADC array of 1024 size that I read into the PICO from ADS1115 through i2c. I am able to convert the values to natural log and that is great, but I have to focus on a specific frequency range (less than 30hz) and I am unsure how to achieve this.

Questions:

  • Is there an equivalent of SCIPY's fftpack.fftfreq() function?
  • Are the FFT values and the index somehow related to the frequency or time? I am wondering if I am missing the obvious here.
  • Will I have to manually calculate these somehow?
  • Is the PICO and CP implementation limited as far as FFT functionality?
2
  • 1
    Perhaps ask ULAB?
    – joan
    Commented Mar 31, 2021 at 7:54
  • Did you get any answers to this?
    – Liam
    Commented Dec 13, 2023 at 16:28

0

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.