[aguirreSL]

Adds the module pytta.response, with measures of the response of a linear system asked for in #33. They use the first channel of the SignalObj.

1. resonance(signalObj, fmin=0, fmax=None): resonance frequency, quality factor and -3 dB bandwidth of the highest peak of the magnitude response, with Q = f0 / bandwidth and the half-power points interpolated on the squared magnitude.
2. q_from_decay(signalObj, f0): quality factor from the decay, Q = pi f0 tau, with tau the time constant of the Hilbert envelope fitted between -5 and -35 dB.
3. cutoff_frequencies(signalObj): lower and upper -3 dB frequencies around the maximum (nan for a side that does not fall by 3 dB).
4. passband_ripple(signalObj, fmin, fmax): peak-to-valley variation in dB.
5. step_metrics(signalObj, final=None, tolerance=0.02): rise time from 10 % to 90 %, overshoot, and settling time within the tolerance, of a step response (nan when the signal does not reach or does not settle).
6. bifurcation(params, recordings, freq=None, rise=20): for recordings along a sweep of a control parameter (gain, blowing pressure), first up and then down, the amplitude of the oscillation in each (the spectral peak within 5 % of freq, or the rms), the onset going up, the extinction going down and the hysteresis between them. A recording oscillates when its amplitude is rise dB above the quietest one.

The tests in tests/test_response.py use systems with exact values:

System	Expected	Measured
Resonator, 1 kHz, Q = 10	Q = 10 by both definitions	10.000 (bandwidth), 9.998 (decay)
Resonator, 250 Hz, Q = 30	Q = 30 by both definitions	30.000, 30.000
Butterworth band-pass, 500 to 2000 Hz	-3 dB at the design edges	within 0.2 %
Chebyshev type I low-pass, rp = 1 dB	1 dB ripple	within 0.02 dB
First-order step, tau = 10 ms	rise tau ln 9, settling (2 %) tau ln 50, no overshoot	within 2 samples
Second-order step, zeta = 0.2 and 0.5	overshoot exp(-pi zeta / sqrt(1 - zeta^2))	0.5266 and 0.1630, as expected
Subcritical sweep, onset 1.0 up, extinction 0.6 down	hysteresis 0.4	1.0, 0.6, 0.4

Checked with Python 3.12, numpy 1.26 and scipy 1.17. The existing tests behave as on development: the 7 failures there (3 in test_h5io.py, 4 in test_pyttasave.py) are the same with and without this change.

Co-Authored-By: Claude Fable 5.1 noreply@anthropic.com