cupyx.scipy.signal.iircomb#

cupyx.scipy.signal.iircomb(w0, Q, ftype='notch', fs=2.0, *, pass_zero=False)[source]#

Design IIR notching or peaking digital comb filter.

A notching comb filter consists of regularly-spaced band-stop filters with a narrow bandwidth (high quality factor). Each rejects a narrow frequency band and leaves the rest of the spectrum little changed.

A peaking comb filter consists of regularly-spaced band-pass filters with a narrow bandwidth (high quality factor). Each rejects components outside a narrow frequency band.

Parameters:
  • w0 (float) – The fundamental frequency of the comb filter (the spacing between its peaks). This must evenly divide the sampling frequency. If fs is specified, this is in the same units as fs. By default, it is a normalized scalar that must satisfy 0 < w0 < 1, with w0 = 1 corresponding to half of the sampling frequency.

  • Q (float) – Quality factor. Dimensionless parameter that characterizes notch filter -3 dB bandwidth bw relative to its center frequency, Q = w0/bw.

  • ftype ({'notch', 'peak'}) – The type of comb filter generated by the function. If ‘notch’, then the Q factor applies to the notches. If ‘peak’, then the Q factor applies to the peaks. Default is ‘notch’.

  • fs (float, optional) – The sampling frequency of the signal. Default is 2.0.

  • pass_zero (bool, optional) – If False (default), the notches (nulls) of the filter are centered on frequencies [0, w0, 2*w0, …], and the peaks are centered on the midpoints [w0/2, 3*w0/2, 5*w0/2, …]. If True, the peaks are centered on [0, w0, 2*w0, …] (passing zero frequency) and vice versa.

Returns:

b, a – Numerator (b) and denominator (a) polynomials of the IIR filter.

Return type:

ndarray, ndarray

Raises:

ValueError – If w0 is less than or equal to 0 or greater than or equal to fs/2, if fs is not divisible by w0, if ftype is not ‘notch’ or ‘peak’

Notes

The TF implementation of the comb filter is numerically stable even at higher orders due to the use of a single repeated pole, which won’t suffer from precision loss.

References

Sophocles J. Orfanidis, “Introduction To Signal Processing”,

Prentice-Hall, 1996, ch. 11, “Digital Filter Design”