Note

This page is a reference documentation. It only explains the class signature, and not how to use it. Please refer to the user guide for the big picture.

nilearn.glm.Contrast#

class nilearn.glm.Contrast(effect, variance, dim=None, dof=10000000000.0, stat_type='t', tiny=1e-50, dofmax=10000000000.0)[source]#

The contrast class handles the estimation of statistical contrasts on a given model: student (t) or Fisher (F).

The important feature is that it supports addition, thus opening the possibility of fixed-effects models.

The current implementation is meant to be simple, and could be enhanced in the future on the computational side (high-dimensional F contrasts may lead to memory breakage).

__init__(effect, variance, dim=None, dof=10000000000.0, stat_type='t', tiny=1e-50, dofmax=10000000000.0)[source]#

Construct instance.

Parameters:

effectarray of shape (contrast_dim, n_voxels): The effects related to the contrast.
variancearray of shape (n_voxels): The associated variance estimate.
dimint or None, optional: The dimension of the contrast.
dofscalar, default=DEF_DOFMAX: The degrees of freedom of the residuals.
stat_type{‘t’, ‘F’}, default=’t’: Specification of the contrast type.
contrast_type: Deprecated since version 0.10.3: Use stat_type instead (see above).
tinyfloat, default=DEF_TINY: Small quantity used to avoid numerical underflows.
dofmaxscalar, default=DEF_DOFMAX: The maximum degrees of freedom of the residuals.

property contrast_type#: Return value of stat_type.

Deprecated since version 0.10.3.

effect_size()[source]#: Make access to summary statistics more straightforward when computing contrasts.

effect_variance()[source]#: Make access to summary statistics more straightforward when computing contrasts.

stat(baseline=0.0)[source]#

Return the decision statistic associated with the test of the null hypothesis: (H0) ‘contrast equals baseline’.

Parameters:

baselinefloat, default=0.0: Baseline value for the test statistic.

Returns:

stat1-d array, shape=(n_voxels,): statistical values, one per voxel.

p_value(baseline=0.0)[source]#

Return a parametric estimate of the p-value associated with the null hypothesis (H0): ‘contrast equals baseline’, using the survival function.

Parameters:

baselinefloat, default=0.0: Baseline value for the test statistic.

Returns:

p_values1-d array, shape=(n_voxels,): p-values, one per voxel

one_minus_pvalue(baseline=0.0)[source]#

Return a parametric estimate of the 1 - p-value associated with the null hypothesis (H0): ‘contrast equals baseline’, using the cumulative distribution function, to ensure numerical stability.

Parameters:

baselinefloat, default=0.0: Baseline value for the test statistic.

Returns:

one_minus_pvalues1-d array, shape=(n_voxels,): one_minus_pvalues, one per voxel

z_score(baseline=0.0)[source]#

Return a parametric estimation of the z-score associated with the null hypothesis: (H0) ‘contrast equals baseline’.

Parameters:

baselinefloat, optional, default=0.0: Baseline value for the test statistic.

Returns:

z_score1-d array, shape=(n_voxels,): statistical values, one per voxel

Examples using `nilearn.glm.Contrast`#

Example of surface-based first-level analysis

Surface-based dataset first and second level analysis of a dataset

nilearn.glm.Contrast#

Examples using nilearn.glm.Contrast#

Examples using `nilearn.glm.Contrast`#