cedalion.models.glm package

Submodules

cedalion.models.glm.basis_functions module

class cedalion.models.glm.basis_functions.Gamma(tau, sigma, T)

Bases: TemporalBasisFunction

Modified gamma function, optionally convolved with a square-wave.

Parameters:

  • tau (Union[Quantity, dict[str, Quantity]]) – onset time

  • sigma (Union[Quantity, dict[str, Quantity]]) – width of the HRF

  • T (Union[Quantity, dict[str, Quantity]]) – convolution width

class cedalion.models.glm.basis_functions.GaussianKernels(t_pre, t_post, t_delta, t_std)

Bases: TemporalBasisFunction

A consecutive sequence of gaussian functions.

Parameters:

  • t_pre (Quantity, [time]) – time before trial onset

  • t_post (Quantity, [time]) – time after trial onset

  • t_delta (Quantity, [time]) – the temporal spacing between consecutive gaussians

  • t_std (Quantiantity, [time]) – time width of the gaussians

class cedalion.models.glm.basis_functions.TemporalBasisFunction(convolve_over_duration)

Bases: ABC

cedalion.models.glm.basis_functions.generate_component_names(n_components)
Return type:

list[str]

cedalion.models.glm.basis_functions.get_other_dim(ts)

Returns the name of 3rd dimension (chromo, wavelength) and it coords.

cedalion.models.glm.basis_functions.to_dict(param, keys)

Extends a parameter to a dict.

cedalion.models.glm.basis_functions.to_unit(obj, unit)

Sets the unit of this quantity or the quantity values in the dict.

cedalion.models.glm.design_matrix module

cedalion.models.glm.design_matrix.build_stim_array(time, onsets, duration)

Build an array indicating active stimulus periods.

The array values are 1 between onset and onset+duration and zero everywhere else.

Parameters:

  • time (Union[_SupportsArray[dtype[Any]], _NestedSequence[_SupportsArray[dtype[Any]]], bool, int, float, complex, str, bytes, _NestedSequence[Union[bool, int, float, complex, str, bytes]]]) – the time axis

  • onsets (Union[_SupportsArray[dtype[Any]], _NestedSequence[_SupportsArray[dtype[Any]]], bool, int, float, complex, str, bytes, _NestedSequence[Union[bool, int, float, complex, str, bytes]]]) – times of stimulus onsets

  • duration (Union[None, _SupportsArray[dtype[Any]], _NestedSequence[_SupportsArray[dtype[Any]]], bool, int, float, complex, str, bytes, _NestedSequence[Union[bool, int, float, complex, str, bytes]]]) – either durations of each stimulus or None, in which case the stimulus duration is set to one sample.

Return type:

ndarray

Returns:

The array denoting

cedalion.models.glm.design_matrix.closest_short_channel(ts_long, ts_short, geo3d)

Create channel-wise regressors use closest nearby short channel.

Parameters:

  • ts_long (DataArray) – time series of long channels

  • ts_short (DataArray) – time series of short channels

  • geo3d (DataArray) – probe geometry

Returns:

channel-wise regressor

cedalion.models.glm.design_matrix.construct_afni_gamma_basis(t_hrf, params_basis)
cedalion.models.glm.design_matrix.construct_basis_functions(t_hrf, idx_basis, params_basis)
cedalion.models.glm.design_matrix.construct_gaussian_basis(t_hrf, params_basis)
cedalion.models.glm.design_matrix.construct_individual_basis(t_hrf, params_basis)
cedalion.models.glm.design_matrix.construct_modified_gamma_basis(t_hrf, params_basis)
cedalion.models.glm.design_matrix.construct_modified_gamma_deriv_basis(t_hrf, params_basis)
cedalion.models.glm.design_matrix.get_ss_regressors(y, geo3d, ss_method='closest', ss_tresh=1.5, as_xarray=False)

Get short separation channels for each long channel.

Parameters:

  • y (DataArray) – xarray.DataArray of the data (time x chromo x channels)

  • geo3d (DataArray) – xarray.DataArray of the 3D geometry (no. of sources/detectors x dim pos)

  • ss_method – method for determining short separation channels (“nearest”, “corr”)

  • ss_tresh (float) – threshold for short separation channels (in cm)

  • get_data – whether to return the short channel data (True) or the short channel names (False)

  • as_xarray (bool) – TBD

Returns:

xarray.DataArray of short separation channels (channels x chromo)

Return type:

ss

cedalion.models.glm.design_matrix.get_third_dimension(ts)
cedalion.models.glm.design_matrix.make_design_matrix(ts_long, ts_short, stim, geo3d, basis_function, drift_order, short_channel_method)

Generate the design matrix for the GLM.

Parameters:

  • ts_long (DataArray) – time series of long distance channels

  • ts_short (Optional[DataArray]) – time series of short distance channels

  • stim (DataFrame) – stimulus DataFrame

  • geo3d (DataArray) – probe geometry

  • basis_function (TemporalBasisFunction) – the temporal basis function(s) to model the HRF

  • drift_order (int | None) – if not None specify the highest polynomial order of the drift terms

  • short_channel_method (str | None) – can be ‘closest’ or ‘max_corr’ and specifies the method to add short channel information ot the design matrix

Returns:

A tuple containing the global design_matrix and a list of channel-wise regressors.

cedalion.models.glm.design_matrix.make_drift_regressors(ts, drift_order)
Return type:

DataArray

cedalion.models.glm.design_matrix.make_hrf_regressors(ts, stim, basis_function)

Create regressors modelling the hemodynamic response to stimuli.

cedalion.models.glm.design_matrix.max_corr_short_channel(ts_long, ts_short)

Create channel-wise regressors using the most correlated short channels.

For each long channel the short channel is selected that has the highest correleation coefficient in any wavelength or chromophore.

Parameters:

  • ts_long (DataArray) – time series of long channels

  • ts_short (DataArray) – time series of short channels

Returns:

channel-wise regressors

cedalion.models.glm.design_matrix.pad_time_axis(time, onsets)

cedalion.models.glm.solve module

cedalion.models.glm.solve.fit(ts, design_matrix, channel_wise_regressors=None, noise_model='ols')

Fit design matrix to data.

Parameters:

  • ts (DataArray) – the time series to be modeled

  • design_matrix (DataArray) – DataArray with dims time, regressor, chromo

  • channel_wise_regressors (list[DataArray] | None) – optional list of design matrices, with additional channel dimension

  • noise_model – must be ‘ols’ for the moment

Returns:

thetas as a DataArray

cedalion.models.glm.solve.hash_channel_wise_regressor(regressor)

Hashes each channel slice of the regressor array.

Parameters:

regressor (DataArray) – array of channel-wise regressors. Dims (channel, regressor, time, chromo|wavelength)

Return type:

list[int]

Returns:

A list of hash values, one hash for each channel.

cedalion.models.glm.solve.iter_design_matrix(ts, design_matrix, channel_wise_regressors=None, channel_groups=None)
cedalion.models.glm.solve.predict(ts, thetas, design_matrix, channel_wise_regressors=None)
Return type:

DataArray

Module contents

Tools for describing fNIRS data with general linear models.