modalities.fmri.design_matrix

Module: modalities.fmri.design_matrix

Inheritance diagram for nipy.modalities.fmri.design_matrix:

This module implements fMRI Design Matrix creation.

The DesignMatrix object is just a container that represents the design matrix. Computations of the different parts of the design matrix are confined to the make_dmtx() function, that instantiates the DesignMatrix object. All the remainder are just ancillary functions.

Design matrices contain three different types of regressors:

1. Task-related regressors, that result from the convolution of the experimental paradigm regressors with hemodynamic models

2. User-specified regressors, that represent information available on the data, e.g. motion parameters, physiological data resampled at the acquisition rate, or sinusoidal regressors that model the signal at a frequency of interest.

3. Drift regressors, that represent low_frequency phenomena of no interest in the data; they need to be included to reduce variance estimates.

Author: Bertrand Thirion, 2009-2011

Class

DesignMatrix

class nipy.modalities.fmri.design_matrix.DesignMatrix(matrix, names, frametimes=None)

Bases: object

This is a container for a light-weight class for design matrices

This class is only used to make IO and visualization.

Attributes:

matrix: array of shape (n_scans, n_regressors)

the numerical specification of the matrix.

names: list of len (n_regressors)

the names associated with the columns.

frametimes: array of shape (n_scans), optional

the occurrence time of the matrix rows.

__init__(matrix, names, frametimes=None)

show(rescale=True, ax=None, cmap=None)

Visualization of a design matrix

Parameters:

rescale: bool, optional

rescale columns magnitude for visualization or not.

ax: axis handle, optional

Handle to axis onto which we will draw design matrix.

cmap: colormap, optional

Matplotlib colormap to use, passed to imshow.

Returns:

ax: axis handle

show_contrast(contrast, ax=None, cmap=None)

Plot a contrast for a design matrix.

Parameters:

contrastnp.float64

Array forming contrast with respect to the design matrix.

ax: axis handle, optional

Handle to axis onto which we will draw design matrix.

cmap: colormap, optional

Matplotlib colormap to use, passed to imshow.

Returns:

ax: axis handle

write_csv(path)

write self.matrix as a csv file with appropriate column names

Parameters:

path: string, path of the resulting csv file

Notes

The frametimes are not written

Functions

nipy.modalities.fmri.design_matrix.dmtx_from_csv(path, frametimes=None)

Return a DesignMatrix instance from a csv file

Parameters:

path: string, path of the .csv file

Returns:

A DesignMatrix instance

nipy.modalities.fmri.design_matrix.dmtx_light(frametimes, paradigm=None, hrf_model='canonical', drift_model='cosine', hfcut=128, drift_order=1, fir_delays=[0], add_regs=None, add_reg_names=None, min_onset=-24, path=None)

Make a design matrix while avoiding framework

Parameters:

see make_dmtx, plus

path: string, optional: a path to write the output

Returns:

dmtx array of shape(nreg, nbframes):

the sampled design matrix

names list of strings of len (nreg)

the names of the columns of the design matrix

nipy.modalities.fmri.design_matrix.make_dmtx(frametimes, paradigm=None, hrf_model='canonical', drift_model='cosine', hfcut=128, drift_order=1, fir_delays=[0], add_regs=None, add_reg_names=None, min_onset=-24)

Generate a design matrix from the input parameters

Parameters:

frametimes: array of shape(nbframes), the timing of the scans

paradigm: Paradigm instance, optional

description of the experimental paradigm

hrf_model: string, optional,

that specifies the hemodynamic response function. Can be one of {‘canonical’, ‘canonical with derivative’, ‘fir’, ‘spm’, ‘spm_time’, ‘spm_time_dispersion’}.

drift_model: string, optional

specifies the desired drift model, to be chosen among ‘polynomial’, ‘cosine’, ‘blank’

hfcut: float, optional

cut period of the low-pass filter

drift_order: int, optional

order of the drift model (in case it is polynomial)

fir_delays: array of shape(nb_onsets) or list, optional,

in case of FIR design, yields the array of delays used in the FIR model

add_regs: array of shape(nbframes, naddreg), optional

additional user-supplied regressors

add_reg_names: list of (naddreg) regressor names, optional

if None, while naddreg>0, these will be termed ‘reg_%i’,i=0..naddreg-1

min_onset: float, optional

minimal onset relative to frametimes[0] (in seconds) events that start before frametimes[0] + min_onset are not considered

Returns:

DesignMatrix instance