ffrprep.analysis.compute_pitch_and_conf

ffrprep.analysis.compute_pitch_and_conf(evoked, win_dur=0.04, hop_dur=0.01, fmin=60, fmax=200, strength_thresh=0.15, smooth_k=3)[source]

Compute the pitch (f0) and confidence metrics for an Evoked response.

Parameters:

  • evoked (mne.Evoked) – Evoked object containing averaged M/EEG data.

  • win_dur (float, default=0.040) – Duration of the sliding window (in seconds) used for pitch estimation.

  • hop_dur (float, default=0.010) – Step size (in seconds) for sliding the window across the signal.

  • fmin (float, default=60) – Minimum frequency (in Hz) to consider for pitch estimation.

  • fmax (float, default=200) – Maximum frequency (in Hz) to consider for pitch estimation.

  • strength_thresh (float, default=0.15) – Minimum normalized autocorrelation strength required to consider a pitch estimate valid.

  • smooth_k (int, default=3) – Kernel size for smoothing the pitch estimates (in number of windows).

Returns:

A dictionary with the following keys:

  • 'times' — array of time points corresponding to the center of each window.

  • 'pitch_hz' — array of pitch estimates (in Hz) for each window.

  • 'pitch_hz_smooth' — smoothed pitch estimates (in Hz) for each window.

  • 'peak_strength' — array of normalized autocorrelation strengths for each window.

  • 'conf_rmax' — array of maximum normalized autocorrelation values for each window.

  • 'conf_pnr' — array of pitch-to-noise ratios for each window.

  • 'conf_z' — array of z-scores for the maximum autocorrelation values for each window.

Return type:

dict

Examples

Compute pitch and confidence metrics for an Evoked response with default parameters.

>>> results = compute_pitch_and_conf(evoked)
>>> print(results['times'])
[0.02 0.03 0.04 ... 0.98 0.99 1.00]
>>> print(results['pitch_hz'])
[120.0 125.0 130.0 ... 110.0 115.0 120.0]
>>> print(results['conf_rmax'])
[0.8 0.85 0.9 ... 0.75 0.8 0.82]