pyhiir 1.0.4

Python wrapper for HIIR library

pyhiir

use hiir (https://github.com/unevens/hiir) library in python

HIIR is a DSP (digital signal processing) library in C++ which allows:

• Changing the sampling rate of a signal by a factor of 2
• Obtaining two signals with a pi/2 phase difference (Hilbert Transform)

Pyhiir offers wrappers for the hiir methods in python.

Usage Examples

Hilbert Transformer

from pyhiir.hiir import hiir
from pyhiir.allpass import Hilbert
import matplotlib.pyplot as plt
import numpy as np
from scipy import signal

if __name__=='__main__':
    h = hiir()

    # compute filter coeficients for 40db attenuation 
    # and transition band of 10% of the sampling frequency
    c = h.compute_coefs(40, .1)

    # create a hilbert transformer with allpass chains
    h = Hilbert(c)

    fi, fq = h.get_transfer_function()
    wi, hi = signal.freqz(fi.b, fi.a, fs=2)
    wq, hq = signal.freqz(fq.b, fq.a, fs=2)

    # Plot the frequency response
    plt.plot(wi, (np.angle(hi)), color='r')
    plt.plot(wq, (np.angle(hq)), color='g')
    
    # plt.plot(wi, (np.angle(hi) - np.angle(hq)), color='y')

    plt.ylabel('Phase [rad]', color='r')
    plt.xlabel('Frequency')
    plt.tick_params('y', colors='r')
    plt.show()
    
    plt.plot(wi, np.abs(hi), color='r')
    plt.plot(wq, np.abs(hq), color='g')
    plt.xlabel('Frequency')
    plt.ylabel('Magnitude [db]')
    plt.ylim(0.95, 1.05)
    plt.show()

Decimator Filter:

from pyhiir.hiir import hiir
from pyhiir.allpass import LowPass
import matplotlib.pyplot as plt
import numpy as np
from scipy import signal

if __name__=='__main__':
    h = hiir()

    # compute filter coeficients for 120db attenuation 
    # and transition band of 1% of the sampling frequency
    c = h.compute_coefs(120, .01)

    # create an half-band low pass filter with allpass chains
    f = LowPass(c)

    # Now get the filter transfer function for plotting
    ff = f.get_transfer_function()
    w, h = signal.freqz(ff.b, ff.a)
    plt.grid(True)
    plt.xlabel("Normalized frequency")
    plt.ylabel("Amplitude [dB]")
    plt.plot(w / np.pi, 20 * np.log10(abs(h)), 'b')
    plt.show()