information and dimensions about files on drive - LIMO-EEG-Toolbox/limo_tools GitHub Wiki

1st level analysis using mass univariate approach

In the subject analysis folder, the data, parameters and results are saved

LIMO.mat

The structure that contains all the information related to the data and the model see https://github.com/LIMO-EEG-Toolbox/limo_eeg/wiki/LIMO-files

Yr.mat

the EEG single trial data reorganized to fit X, that is grouped by conditions if Cat ~=0 dimension [channels x frames x trials]

Yhat.mat

the predicted data dimension [channels x frames x trials]

Beta.mat

the beta values (parameter estimates) dimension [channels x frames x number of parameters in the model (columns of X)]

Res.mat

the residuals (non modelled) data dimension [channels x frames x trials]

R2.mat

the model fit statistic, i.e. percentage of variance explained dimension [channels x frames x R2/F/p values]

Condition_effect_X

refers to a factor effect in categorical designs dimension [channels x frames x F/p values)

Interaction_effect_X

refers to an interaction between factors dimension [channels x frames x F/p values)

Covariate_effect_X

refers to the effect of a continuous regressor dimension [channels x frames x F/p values).

semi_partial_coef_X.mat

refers to the semi-partial coefficient of a factor (defined by LIMO.design.nb_conditions) or a covariate (defined in LIMO.design.nb_continuous) dimension [channels x frames x R2/F/p values].

con_X.mat

refers to a t contrast dimension [channels x frames x cB/standard error/df/t/p values]

ess_X.mat

refers to a F contrast dimension [channels x frames x cB/standard error/df/F/p values].


if bootstrap is selected at import, in the newly created H0 subfolder we have similar files computed under H0 with n bootstrap (by default 800):

boot_table

the resampling table used dimension [number of trials x number of bootstraps]

H0_Betas

dimension [channels x frames x number of parameters in the model (columns of X) x number of bootstraps]

H0_R2

dimension: [channels x frames x R2/F/p values x number of bootstraps]

H0_Condition_effect_X

dimension [channels x frames x F/p values x number of bootstraps]

H0_Interaction_effect_X

dimension [channels x frames x F/p values x number of bootstraps]

H0_Covariate_effect_X

dimension [channels x frames x F/p values x number of bootstraps]

H0_semi_partial_coef_X.mat

dimension [channels x frames x R2/F/p values x number of bootstraps].

H0_con_X.mat

dimension [channels x frames x cB/t/p values x number of bootstraps].

H0_ess_X.mat

dimension [channels x frames x cB/F/p values x number of bootstraps].


if tfce is selected at import, we have similar files as above but computed with tfce:

tfce_R2

dimension: [channels x frames]

tfce_Condition_effect_X

dimension [channels x frames]

tfce_Interaction_effect_X

dimension [channels x frames]

tfce_Covariate_effect_X

dimension [channels x frames]

tfce_semi_partial_coef_X.mat

dimension [channels x frames].

tfce_con_X.mat

dimension [channels x frames].

tfce_ess_X.mat

dimension [channels x frames].

In the H0 subfolder, the tfce score maps under H0

tfce_H0_R2

dimension [channels x frames x number of bootstraps]

tfce_H0_Condition_effect_X

dimension [channels x frames x number of bootstraps]

tfce_H0_Interaction_effect_X

dimension [channels x frames x number of bootstraps]

tfce_H0_Covariate_effect_X

dimension [channels x frames x number of bootstraps]

tfce_H0_semi_partial_coef_X.mat

dimension [channels x frames x number of bootstraps].

tfce_H0_con_X.mat

dimension [channels x frames x number of bootstraps].

tfce_H0_ess_X.mat

dimension [channels x frames x number of bootstraps].

2nd level analysis using mass univariate approach

LIMO.mat

The structure that contains all the information related to the data and the model see https://github.com/LIMO-EEG-Toolbox/limo_eeg/wiki/LIMO-files

One sample t-test

Yr.mat

The data used for the analysis.

one_sample_ttest_parameter_X.mat

Returns the trimmed mean parameter values and associated statistics. Dimensions: [electrodes x frames x 5]. The last dimension codes mean values, standard error, degrees of freedom, t and p.

H0_one_sample_ttest_parameter_X.mat

This file contains the T and p values obtained under H0 for each bootstrap. Dimensions: [electrodes x frames x 2 x number of bootstraps]

tfce_one_sample_ttest_parameter_X.mat

tfce scores of the t-test. Dimensions: [electrodes x frames]

H0_tfce_one_sample_ttest_parameter_X.mat

tfce scores of the t–test under H0. Dimensions: [electrodes x frames x number of bootstraps]

Two samples t-test

Yr1.mat and Yr2.mat

This is the data for each group. Dimensions: [electrodes x frames x nb subjects in group]

two_samples_ttest_parameter_X.mat

Returns the parameter values and associated statistics. Dimensions: [electrodes x frames x 5]. The last dimension codes for trimmed mean differences, standard error, degrees of freedom, t and p.

H0_two_samples_ttest_parameter_X.mat

This file contains the trimmed mean differences, t and p values obtained under H0 for each bootstrap. Dimensions: [electrodes x frames x 3 x number of bootstraps]

tfce_two_samples_ttest_parameter_X.mat

tfce scores of the t-test. Dimensions: [electrodes x frames]

H0_tfce_two_samples_ttest_parameter_X.mat

tfce scores of the t–test under H0. Dimensions: [electrodes x frames x number of bootstraps]

Paired t-test

Yr1.mat and Yr2.mat

This is the data for each parameter. Dimensions: [electrodes x frames x subjects]

paired_samples_ttest_parameter_X.mat

Returns the parameter values and associated statistics. Dimensions: [electrodes x frames x 5]. The last dimension codes for trimmed mean difference, df, standard error, t and p.

H0_paired_samples_ttest_parameter_X.mat

This file contains the trimmed mean differences, T and p values obtained under H0 for each bootstrap. Dimensions: [electrodes x frames x 3 x number of bootstraps]

tfce_paired_samples_ttest_paramater_X.mat

tfce scores of the t-test. Dimensions: [electrodes x frames]

H0_tfce_paired_samples_ttest_parameter_X.mat

tfce scores of the t–test under H0. Dimensions: [electrodes x frames x number of bootstraps]

Regression analysis

Same as for 1st level.

R2.mat

The model fit statistic, i.e. percentage of variance explained. Dimensions: [electrodes x frames x R2/F/p values]

H0_R2.mat

Dimensions: [electrodes x frames x R2/F/p values x number of bootstraps]

Covariate_effect_X.mat

Refers to the effect of a continuous regressor. Dimensions: [electrodes x frames x F/p values).

H0_Covariate_effect_X.mat

Dimensions: [electrodes x frames x F/p values x number of bootstraps]

ANOVA/ANCOVA

Same as for 1st level.

Condition_effect_X.mat

refers to a factor effect in categorical designs dimension [channels x frames x F/p values)

H0_Condition_effect_X.mat

dimension [channels x frames x F/p values x number of bootstraps]

Covariate_effect_X.mat

refers to the effect of a continuous regressor dimension [channels x frames x F/p values).

H0_Covariate_effect_X.mat

dimension [channels x frames x F/p values x number of bootstraps]

repeated measure ANOVA

Yr.mat

Dimensions: [electrodes x frames x subjects x rep measures]

Rep_ANOVA_Factor_X.mat

Dimensions: [electrodes x frames x F/p values]

Rep_ANOVA_Gp_effect.mat

Dimensions: [electrodes, frames, F/p values]

Rep_ANOVA_Interaction_gp_Factor_X.mat

Dimensions: [electrodes x frames x F/p values]

H0_XXX same as above including number of bootstraps as last dimension

e.g., H0_Rep_ANOVA_Factor_X.mat Dimensions: [electrodes x frames x F/ p values x number of bootstraps]

tfce_Rep_ANOVA_XXX

tfce_H0_Rep_ANOVA_XXX