spkit.add_noise¶
- spkit.add_noise(x, snr_db=10, return_noise=False)¶
Add Gaussian Noise to Signal
Add Gaussian Noise to Signal
SNR = sig_pw/noise_pw SNR_db = sig_pw_db - noise_pw_db
noise_pw = sig_pw/SNR noise_pw = 10**( (sig_pw_db-SNR_db)/10 )
noise ~ N(0, sqrt(noise_pw))
- Parameters:
- x: nd array
input signal, 1d or multi-dimensional
- snr_db: scalar
Desired Signal to Noise Ratio (SNR) in Decible (dB)
- return_noise: bool, default=False
if True, returns the added noise
- Returns:
- y: array, same shape as x
Noisy signal
Notes
#TODO
Examples
import numpy as np import matplotlib.pyplot as plt import spkit as sp np.random.seed(1) t = np.linspace(0,2,200) x = np.cos(2*np.pi*1*t) SNRs = [30,20,10,5,0] plt.figure(figsize=(10,5)) plt.plot(t,x,label=f'x: signal') for i,snr in enumerate(SNRs): plt.plot(t,sp.add_noise(x,snr)-(i+1)*2,label=f'SNR = {snr} dB') plt.xlim([t[0],t[-1]]) plt.legend(bbox_to_anchor=(1,1)) plt.grid() plt.yticks([]) np.random.seed(None) plt.tight_layout() plt.show()
