How to find Fourier Transform Coefficients from a .wav file?

Question

Well, I have done a good amount of research on Speech data, I was able to return discrete Fourier transform of real or complex sequence from a wav file.

import matplotlib.pyplot as plt
from scipy.io import wavfile as wav
from scipy.fftpack import fft
import numpy as np
rate, data = wav.read('Data/File1.wav')
fft_out = fft(data)
%matplotlib inline
plt.plot(data, np.abs(fft_out))
plt.show()

When I tried to save this np.abs(fft_out) data, it is ultimately a 1D array data, which is quite obvious

But I need some help in extracting features (fourier coefficients) from the fourier transformed data, can somebody show some light?

Ultimately, I am being asked to create a final dataset with important features from fourier transformed signals

Note: I don't need the exact code for this, but any direction would be helpful. My main task is go get important features extracted from a fourier transformed data.

Answer 1

outputs of fft function are fourier coefficients. for each frequency there is one coefficient in 1D array.

Fourier equation

as in this equation cn is coefficient . for x=0 the coefficient is fft_out[0]