bandpass_ndh.py

from scipy.signal import butter, lfilter

def bandpass_ndh(data, times, lowcut, highcut, order=5):
    """
    % (C) Nick Holschuh - Amherst College - 2022 (Nick.Holschuh@gmail.com)
    % This function applies a bandpass filter on an input series.
    %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
    % The inputs are as follows:
    %
    % data -- the input array to be filtered
    % times -- the array defining the axis of variability that describes "data"
    % lowcut -- the frequency that defines the high-pass transition
    % highcut -- the frequency that defines the low-pass transition
    % order -- the order that defines the butterworth filter
    %
    %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
    % The outputs are as follows:
    %
    % y -- the filtered dataset
    %
    %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
    %% Modified from https://stackoverflow.com/questions/12093594/how-to-implement-band-pass-butterworth-filter-with-scipy-signal-butter
    """
    fs = 1/(times[1]-times[0])
    nyq = 0.5 * fs
    low = lowcut / nyq
    high = highcut / nyq
    b, a = butter(order, [low, high], btype='band')
    y = lfilter(b, a, data)
    return y