NumPy and FFT

[Michael A. Miller]

If your data is in an array y, a FFT can be calculated like this:

import FFT
fft = FFT.fft(y)

A fast FT doesn't care at all what the sampling rate is, so
you'll have to calculate your own frequencies.  If your sample
rate is f_sample, you could do something like this:

import Numeric
delta = 1 / f_sample
f = Numeric.array(range(len(fft))) / delta / len(fft)
N = len(f)
for i in range(N/2):
    f[N/2+i] = - result[N/2-i]

Mike

P.S. More NumPy savvy folks can probably point out a much more
efficient way of calculating the negative frequencies...