Spectral¶

What's happening in frequency space?

Features in the cobrabox.spectral domain. Access them as cb.<Feature> (canonical), or as cb.spectral.<Feature> / cb.feature.<Feature>.

Each feature has two forms: a class for building pipelines, and a one-shot function for a single call.

Compute band power for specified frequency bands using Welch's method.

cb.band_power(data, ...)  # one-shot
cb.BandPower(...).apply(data)  # composable, for pipelines

Tags: power-spectrum, frequency-band, welch, alpha, theta, delta, beta, gamma, epilepsy, depression, anesthesia, sleep, eeg, fmri, io:scalar-per-channel-per-band

Continuous wavelet transform (CWT) scalogram.

cb.continuous_wavelet_transform(data, ...)  # one-shot
cb.ContinuousWaveletTransform(...).apply(data)  # composable, for pipelines

Tags: wavelet, scalogram, time-frequency, scale-adaptive, seizure-onset, erps, eeg, io:frequency-time-output

Compute cordance, a qEEG measure combining absolute and relative bandpower.

cb.cordance(data, ...)  # one-shot
cb.Cordance(...).apply(data)  # composable, for pipelines

Tags: qeeg, absolute-power, relative-power, depression, treatment-response, eeg, io:scalar-per-channel-per-band

Multi-level discrete wavelet decomposition (DWT).

cb.discrete_wavelet_transform(data, ...)  # one-shot
cb.DiscreteWaveletTransform(...).apply(data)  # composable, for pipelines

Tags: wavelet, sub-band, dyadic, decomposition, denoising, compression, eeg, io:adds-dimension

Power spectral density via FFT.

cb.power_spectral_density(data, ...)  # one-shot
cb.PowerSpectralDensity(...).apply(data)  # composable, for pipelines

Tags: psd, power-spectrum, frequency-domain, eeg, fmri

Compute the power spectrogram for each spatial channel.

cb.spectrogram(data, ...)  # one-shot
cb.Spectrogram(...).apply(data)  # composable, for pipelines

Tags: stft, time-frequency, power-spectrum, seizure-onset, event-related, eeg, fmri, io:frequency-time-output