me_filter.py

# filter some data using SOS so as to preserve the dynamic range

import numpy as np
import scipy.signal as sig

fs = 1024

bg_raw = np.random.randn(1000)

# This is a bandpass
f1    = 15
f2    = 100
sosBP1 = sig.butter(8, [f1/(fs/2), f2/(fs/2)], btype='bandpass', output='sos')

# invertable band pass
fz = [4, 400]
fp = [15, 20, 150]

_,z0,_ = sig.butter(10, fz[0]/(fs/2), btype='lowpass', output='zpk')
_,p0,_ = sig.butter( 8, fp[0]/(fs/2), btype='lowpass', output='zpk')
_,p1,_ = sig.butter( 1, fp[1]/(fs/2), btype='lowpass', output='zpk')
_,p2,_ = sig.butter( 3, fp[1]/(fs/2), btype='lowpass', output='zpk')
_,z1,_ = sig.butter( 2, fz[1]/(fs/2), btype='lowpass', output='zpk')

zs    = np.concatenate((z0, z1))
ps    = np.concatenate((p0, p1))
BP    = sig.zpk2sos(zs, ps, 1, pairing='keep_odd')
sosBP = np.concatenate([sosBP1, BP], axis = 0)

# filter the data using many second-order-sections
bg    = sig.sosfilt(sosBP,  bg_raw)