Coherence analysis (spynal.sync.coherence)

Oscillatory coherence analysis

coherence(data1, data2, axis=0, return_phase=False, transform=None, single_trial=None, spec_method='wavelet', data_type=None, smp_rate=None, time_axis=None, taper_axis=None, keepdims=True, **kwargs)

Compute coherence between pair of channels of raw or spectral (time-frequency) data (LFP or spikes)

Coherence is a spectral analog of linear correlation that takes both phase and amplitude into account.

Only parameters differing from synchrony() are described here.

Parameters:

  • transform ('Z' or None, default: None) – Transform to apply to all computed coherence values. Set=None to return raw, untransformed coherence. Set=’Z’ to Z-transform coherence using Jarvis & Mitra (2001) method.

  • **kwargs – Any other keyword args passed as-is to spectrogram() function.

spike_field_coherence(spkdata, lfpdata, axis=0, time_axis=None, taper_axis=None, timepts=None, transform=None, data_type=None, spec_method='multitaper', smp_rate=None, return_phase=False, keepdims=True, **kwargs)

Compute pairwise coherence between single-channel spiking data and LFP data

Only parameters differing from spike_field_coupling() are described here.

Parameters:

  • transform ('Z' or None, default: None) – Transform to apply to all computed coherence values. Set=None to return raw, untransformed coherence. Set=’Z’ to Z-transform coherence using Jarvis & Mitra (2001) method.

  • **kwargs – Any other keyword args passed as-is to spectrogram() function

ztransform_coherence(coh, df, beta=1.15)

z-transform coherence values to render them approximately normally distributed

Parameters:

  • coh (ndarray, shape=Any) – Raw coherence values

  • df (int) – Degrees of freedom of coherence estimates. For multitaper spectral estimates, this is usually df = 2*n_trials*n_tapers. For other estimates, this is usually df = 2*n_trials.

  • beta (scalar, default: 23/20) – Mysterious number from Jarvis & Mitra to make output z-scores approximately normal

Returns:

z – z-transformed coherence. Same shape as coh.

Return type:

ndarray, shape=Any

References