scripting 1st level analysis

Scripting and batching many subjects (limo_batch)

This is the preferred option, as this calls limo_batch and the psom pipeline engine which captures error and thus avoid crashing even if one subject fails - and output a report.

Use EEGLAB STUDY

If you have a STUDY in EEGLAB, then all is needed is define the design, i.e. which variables to use, then call LIMO. This can be done in two (actually 3 if we count the save) commands: std_makedesign and pop_limo.

For instance doing:

STUDY = std_makedesign(STUDY, ALLEEG, 1, 'name','FaceRepetition','delfiles','off','defaultdesign','off','variable1','type','values1',{'famous_new','famous_second_early','famous_second_late','scrambled_new','scrambled_second_early','scrambled_second_late','unfamiliar_new','unfamiliar_second_early','unfamiliar_second_late'}, 'vartype1','categorical','subjselect',{'sub-002','sub-003','sub-004','sub-005','sub-006','sub-007','sub-008','sub-009','sub-010','sub-011','sub-012','sub-013','sub-014','sub-015','sub-016','sub-017','sub-018','sub-019'});
[STUDY, EEG] = pop_savestudy( STUDY, EEG, 'savemode','resave');
STUDY = pop_limo(STUDY, ALLEEG, 'method','OLS','measure','daterp','timelim',[-50 650],'erase','on','splitreg','off','interaction','off');

one creates a design (here with 9 conditions) and a given set of subjects using std_makedesign, saves it to make sure all fields are up to date, and call pop_limo with LIMO options.

Define the model yourself and call limo_batch

start by indicating files to load

model.set_files = {'D:\DataSets\Face_repetition\sub-02\eeg\sub-02_proc.set'}; % cell array of subjects  
model.cat_files = {'D:\DataSets\Face_repetition\sub-02\eeg\cat_factorial_fake.txt'};  
model.cont_files = []; % a cell array of continuous variable or variable files  

then indicates the model parameters

model.defaults.type = 'Channels';  
model.defaults.analysis = 'Time';  
model.defaults.method = 'WLS'; % 'OLS'  
model.defaults.type_of_analysis = 'univariate';  
model.defaults.fullfactorial = 1;  
model.defaults.zscore = 0;  
model.defaults.start = -50; % starting time in ms  
model.defaults.end = 600; % ending time in ms  
model.defaults.bootstrap = 0;  
model.defaults.tfce = 0;  
neighbours = load('D:\DataSets\Face_repetition\derivatives\limo_gp_level_chanlocs.mat')  
model.defaults.neighbouring_matrix = neighbours.channeighbstructmat;  

finally run limo_batch.m

[LIMO_files, procstatus] = limo_batch('model specification',model,[])

Add contrasts

Using limo_batch is the easiest option, for instance

contrast.LIMO_files      = all_my_LIMO.mat_files_as_list; 
contrast.mat             = [1 1 1 -1 -1 -1 0 0 0 0 ; 0 0 0 1 1 1 -1 -1 -1 0];
confiles                 = limo_batch('contrast only',[],contrast,STUDY); 

will compute 3 contrasts for all subjects listed in contrast.LIMO_files

Scripting per subject (making LIMO.mat)

Below is an example of code that will run the analysis for one subject, this does the same job as calling limo_batch above but gives you access to lower level functions and can thus be adapted for each subject if needed.

% load EEGLAB data  
EEGLIMO                       =  pop_loadset('D:\DataSets\Face_repetition\sub-02\eeg\sub-02_proc.set');  

% set fields for data  
LIMO.data.data_dir            = 'D:\DataSets\Face_repetition\sub-02\eeg';  
LIMO.data.data                = 'sub-02_proc.set';  
LIMO.data.chanlocs            = EEGLIMO.chanlocs;  
LIMO.data.sampling_rate       = EEGLIMO.srate;  
LIMO.data.Cat                 = load('cat_factorial_fake.txt');  
LIMO.data.Cont                = [];  
LIMO.data.start               = -50;  
LIMO.data.trim1               = 476;  
LIMO.data.end                 = 600;  
LIMO.data.trim2               = 801;  
LIMO.data.timevect            = EEGLIMO.times(LIMO.data.trim1:LIMO.data.trim2);  
    
% set design`
LIMO.design.fullfactorial     = 1;  
LIMO.design.zscore            = 0;  
LIMO.design.method            = 'WLS';  
LIMO.design.type_of_analysis  = 'Mass-univariate';  
LIMO.design.bootstrap         = 0;  
LIMO.design.tfce              = 0;  
    
% LIMO stuff `  
LIMO.Type                     = 'Channels';  
LIMO.Level                    = 1;  
LIMO.Analysis                 = 'Time';  
LIMO.dir                      = 'D:\DataSets\Face_repetition\sub-02\eeg\WLS';  

mkdir('D:\DataSets\Face_repetition\sub-02\eeg\WLS')  
cd('D:\DataSets\Face_repetition\sub-02\eeg\WLS')  
save('D:\DataSets\Face_repetition\sub-02\eeg\WLS\LIMO.mat','LIMO')  

% evaluate the design  
[LIMO.design.X,LIMO.design.nb_conditions,LIMO.design.nb_interactions,LIMO.design.nb_continuous] = ...  
    limo_design_matrix(EEGLIMO.data(:,LIMO.data.trim1:LIMO.data.trim2,:),LIMO,1);  
LIMO.design.status = 'to do';  
save('D:\DataSets\Face_repetition\sub-02\eeg\WLS\LIMO.mat','LIMO')  

% run the GLM 
limo_eeg(4,pwd)