I am using

PSD = Abs[Fourier[normedAutoValues, FourierParameters -> {-1, 1}]];

normedAutoValues contains around 200 *10^7 data points.

So it takes for ever... Is there a way to monitor the progress of Fourier[]?

  • $\begingroup$ If you have so much data on hand, you might want to look into implementing a pruned FFT. $\endgroup$ – J. M.'s technical difficulties May 31 '15 at 11:22
  • $\begingroup$ You might also consider using Spectrogram, which nicely partitions the data and takes the spectra in reasonably sized chunks. If what you want is the PSD (power spectral density), as suggested by your variable name, then a single FFT is not the way to do it. You might want to consider PowerSpectralDensity . $\endgroup$ – bill s May 31 '15 at 11:46
  • $\begingroup$ In my case it is the PSD, because the data is already the autocorrelation of the signal. Thanks for the tip with Spectrogram. I will try that. $\endgroup$ – newandlost May 31 '15 at 12:34
  • $\begingroup$ Why do you need so much points? How about resampling the data? Do you really need to capture such high components of the FFT? $\endgroup$ – Bichoy Jun 1 '15 at 5:11

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.