From 1ab0863a84b433e9aa9ac3d83f1530ca1475e22e Mon Sep 17 00:00:00 2001
From: Felix Zakirov <fzaki001@fiu.edu>
Date: Sun, 8 Dec 2024 23:39:45 -0500
Subject: [PATCH] Updated scripts
---
code/preprocessing-eeg/create_mat.m | 63 ++++++-------
code/preprocessing-eeg/create_mat_csd.m | 91 +++++++------------
.../tf/MainScript_Calculate_TF_ITPS_ICPS.m | 38 +++++---
.../MainScript_Calculate_TF_ITPS_ICPS_para.m | 35 ++++---
code/preprocessing-eeg/tf/timefreq_script.m | 5 +-
code/preprocessing-eeg/transform_csd.m | 70 ++++++++------
6 files changed, 152 insertions(+), 150 deletions(-)
diff --git a/code/preprocessing-eeg/create_mat.m b/code/preprocessing-eeg/create_mat.m
index 660f442..42766d1 100644
--- a/code/preprocessing-eeg/create_mat.m
+++ b/code/preprocessing-eeg/create_mat.m
@@ -14,6 +14,8 @@
% addpath(genpath([main_dir filesep 'MADE-EEG-preprocessing-pipeline']));% enter the path of the EEGLAB folder in this line
addpath(genpath('/home/data/NDClab/tools/lab-devOps/scripts/MADE_pipeline_standard/eeg_preprocessing'));% enter the path of the folder in this line
+addpath(genpath('/home/data/NDClab/analyses/thrive-theta-ddm/code/matlab/'))
+
%Location of "EEG
% addpath(genpath([main_dir filesep 'eeglab13_4_4b']));% enter the path of the EEGLAB folder in this line
addpath(genpath('/home/data/NDClab/tools/lab-devOps/scripts/MADE_pipeline_standard/eeglab13_4_4b'));% enter the path of the EEGLAB folder in this line
@@ -30,7 +32,7 @@
%location of dataset folder
% dataset_dir = '/Users/fzaki001/thrive-theta-ddm';
-dataset_dir = '/home/data/NDClab/datasets/thrive-dataset/derivatives/preprocessed/';
+dataset_dir = '/home/data/NDClab/datasets/thrive-dataset';
% summary_csv_path = '/Users/fzaki001/thrive-theta-ddm/derivatives/behavior/summary.csv';
summary_csv_path = '/home/data/NDClab/analyses/thrive-theta-ddm/derivatives/behavior/summary.csv';
@@ -81,7 +83,10 @@
% outputHeader = {'id, s_resp_incon_error, s_resp_incon_corr, ns_resp_incon_error, ns_resp_incon_corr'};
% dlmwrite(strcat('thrive_trialCounts_respOnly', date, '.csv'), outputHeader, 'delimiter', '', '-append');
+diary(sprintf('erp_log_%s.log', datestr(now, 'mm_dd_yyyy_HH_MM_SS')))
+
for subject=1:length(datafile_names)
+
EEG=[];
fprintf('\n\n\n*** Processing subject %d (%s) ***\n\n\n', subject, datafile_names{subject});
@@ -138,14 +143,14 @@
%count how many of each event type (combination of event types) of
%interest are present
- s_resp_incon_error = length(find( (strcmp({EEG.event.observation}, "s")) & (strcmp({EEG.event.eventType}, "resp")) & (strcmp({EEG.event.congruency}, "i")) & ([EEG.event.accuracy] == 0) & ([EEG.event.responded] == 1) & ([EEG.event.validRt] == 1) ));
- s_resp_incon_corr = length(find( (strcmp({EEG.event.observation}, "s")) & (strcmp({EEG.event.eventType}, "resp")) & (strcmp({EEG.event.congruency}, "i")) & ([EEG.event.accuracy] == 1) & ([EEG.event.responded] == 1) & ([EEG.event.validRt] == 1) ));
- ns_resp_incon_error = length(find( (strcmp({EEG.event.observation}, "ns")) & (strcmp({EEG.event.eventType}, "resp")) & (strcmp({EEG.event.congruency}, "i")) & ([EEG.event.accuracy] == 0) & ([EEG.event.responded] == 1) & ([EEG.event.validRt] == 1) ));
- ns_resp_incon_corr = length(find( (strcmp({EEG.event.observation}, "ns")) & (strcmp({EEG.event.eventType}, "resp")) & (strcmp({EEG.event.congruency}, "i")) & ([EEG.event.accuracy] == 1) & ([EEG.event.responded] == 1) & ([EEG.event.validRt] == 1) ));
- s_stim_incon_corr = length(find( (strcmp({EEG.event.observation}, "s")) & (strcmp({EEG.event.eventType}, "stim")) & (strcmp({EEG.event.congruency}, "i")) & ([EEG.event.accuracy] == 1) & ([EEG.event.responded] == 1) & ([EEG.event.validRt] == 1) ));
- s_stim_con_corr = length(find( (strcmp({EEG.event.observation}, "s")) & (strcmp({EEG.event.eventType}, "stim")) & (strcmp({EEG.event.congruency}, "c")) & ([EEG.event.accuracy] == 1) & ([EEG.event.responded] == 1) & ([EEG.event.validRt] == 1) ));
- ns_stim_incon_corr = length(find( (strcmp({EEG.event.observation}, "ns")) & (strcmp({EEG.event.eventType}, "stim")) & (strcmp({EEG.event.congruency}, "i")) & ([EEG.event.accuracy] == 1) & ([EEG.event.responded] == 1) & ([EEG.event.validRt] == 1) ));
- ns_stim_con_corr = length(find( (strcmp({EEG.event.observation}, "ns")) & (strcmp({EEG.event.eventType}, "stim")) & (strcmp({EEG.event.congruency}, "c")) & ([EEG.event.accuracy] == 1) & ([EEG.event.responded] == 1) & ([EEG.event.validRt] == 1) ));
+ s_resp_incon_error = length(find( (strcmp({EEG.event.observation}, "s")) & (strcmp({EEG.event.eventType}, "resp")) & (strcmp({EEG.event.congruency}, "i")) & ([EEG.event.accuracy] == 0) & ([EEG.event.responded] == 1) & ([EEG.event.validRt] == 1) & ([EEG.event.extraResponse] == 0) ));
+ s_resp_incon_corr = length(find( (strcmp({EEG.event.observation}, "s")) & (strcmp({EEG.event.eventType}, "resp")) & (strcmp({EEG.event.congruency}, "i")) & ([EEG.event.accuracy] == 1) & ([EEG.event.responded] == 1) & ([EEG.event.validRt] == 1) & ([EEG.event.extraResponse] == 0) ));
+ ns_resp_incon_error = length(find( (strcmp({EEG.event.observation}, "ns")) & (strcmp({EEG.event.eventType}, "resp")) & (strcmp({EEG.event.congruency}, "i")) & ([EEG.event.accuracy] == 0) & ([EEG.event.responded] == 1) & ([EEG.event.validRt] == 1) & ([EEG.event.extraResponse] == 0) ));
+ ns_resp_incon_corr = length(find( (strcmp({EEG.event.observation}, "ns")) & (strcmp({EEG.event.eventType}, "resp")) & (strcmp({EEG.event.congruency}, "i")) & ([EEG.event.accuracy] == 1) & ([EEG.event.responded] == 1) & ([EEG.event.validRt] == 1) & ([EEG.event.extraResponse] == 0) ));
+ s_stim_incon_corr = length(find( (strcmp({EEG.event.observation}, "s")) & (strcmp({EEG.event.eventType}, "stim")) & (strcmp({EEG.event.congruency}, "i")) & ([EEG.event.accuracy] == 1) & ([EEG.event.responded] == 1) & ([EEG.event.validRt] == 1) & ([EEG.event.extraResponse] == 0) ));
+ s_stim_con_corr = length(find( (strcmp({EEG.event.observation}, "s")) & (strcmp({EEG.event.eventType}, "stim")) & (strcmp({EEG.event.congruency}, "c")) & ([EEG.event.accuracy] == 1) & ([EEG.event.responded] == 1) & ([EEG.event.validRt] == 1) & ([EEG.event.extraResponse] == 0) ));
+ ns_stim_incon_corr = length(find( (strcmp({EEG.event.observation}, "ns")) & (strcmp({EEG.event.eventType}, "stim")) & (strcmp({EEG.event.congruency}, "i")) & ([EEG.event.accuracy] == 1) & ([EEG.event.responded] == 1) & ([EEG.event.validRt] == 1) & ([EEG.event.extraResponse] == 0) ));
+ ns_stim_con_corr = length(find( (strcmp({EEG.event.observation}, "ns")) & (strcmp({EEG.event.eventType}, "stim")) & (strcmp({EEG.event.congruency}, "c")) & ([EEG.event.accuracy] == 1) & ([EEG.event.responded] == 1) & ([EEG.event.validRt] == 1) & ([EEG.event.extraResponse] == 0) ));
%Create the trial counts table for trial counts
counts_table=table({datafile_names{subject}}, {s_resp_incon_error}, {s_resp_incon_corr}, {ns_resp_incon_error}, {ns_resp_incon_corr}, {s_stim_incon_corr}, {s_stim_con_corr}, {ns_stim_incon_corr}, {ns_stim_con_corr});
@@ -165,7 +170,7 @@
%specify min number of trials per condition (if file contains less than
%this number for ANY condition, then they will be skipped for ALL conditions
-minTrials = 8;
+minTrials = 6;
%specify min accuracy per condition (if file contains less than
%this number for ANY condition, then they will be skipped for ALL conditions
@@ -208,32 +213,18 @@
EEG = pop_selectevent( EEG, 'latency','-.1 <= .1','deleteevents','on');
EEG = eeg_checkset( EEG );
- % NOTE %
- %the logic of checking conditions and then looping over conditions
- %below is fairly hard-coded and could be be substantially improved to
- %allow for easier reuse when number of conditions or number of
- %variables per condition changes.
- %
- %before pulling trials of interest, for any conditions, check to make
- %sure this file/participant has more than minTrials for EACH condition.
- %If the file/participant is below minTrials for even one of the
- %conditions that will be pulled, then the file/participant is skipped
- %entirely and no condition data at all will be pulled for this specific
- %file (but the participant can still have data pulled for another one
- %of their files from another visit).
- %
%count trials for each condition of interest and store in numTrials vector
- numTrials(1) = length(find( (strcmp({EEG.event.observation}, "s")) & (strcmp({EEG.event.eventType}, "resp")) & (strcmp({EEG.event.congruency}, "i")) & ([EEG.event.accuracy] == 0) & ([EEG.event.responded] == 1) & ([EEG.event.validRt] == 1) ));
- numTrials(2) = length(find( (strcmp({EEG.event.observation}, "s")) & (strcmp({EEG.event.eventType}, "resp")) & (strcmp({EEG.event.congruency}, "i")) & ([EEG.event.accuracy] == 1) & ([EEG.event.responded] == 1) & ([EEG.event.validRt] == 1) ));
- numTrials(3) = length(find( (strcmp({EEG.event.observation}, "ns")) & (strcmp({EEG.event.eventType}, "resp")) & (strcmp({EEG.event.congruency}, "i")) & ([EEG.event.accuracy] == 0) & ([EEG.event.responded] == 1) & ([EEG.event.validRt] == 1) ));
- numTrials(4) = length(find( (strcmp({EEG.event.observation}, "ns")) & (strcmp({EEG.event.eventType}, "resp")) & (strcmp({EEG.event.congruency}, "i")) & ([EEG.event.accuracy] == 1) & ([EEG.event.responded] == 1) & ([EEG.event.validRt] == 1) ));
+ numTrials(1) = length(find( (strcmp({EEG.event.observation}, "s")) & (strcmp({EEG.event.eventType}, "resp")) & (strcmp({EEG.event.congruency}, "i")) & ([EEG.event.accuracy] == 0) & ([EEG.event.responded] == 1) & ([EEG.event.validRt] == 1) & ([EEG.event.extraResponse] == 0) ));
+ numTrials(2) = length(find( (strcmp({EEG.event.observation}, "s")) & (strcmp({EEG.event.eventType}, "resp")) & (strcmp({EEG.event.congruency}, "i")) & ([EEG.event.accuracy] == 1) & ([EEG.event.responded] == 1) & ([EEG.event.validRt] == 1) & ([EEG.event.extraResponse] == 0) ));
+ numTrials(3) = length(find( (strcmp({EEG.event.observation}, "ns")) & (strcmp({EEG.event.eventType}, "resp")) & (strcmp({EEG.event.congruency}, "i")) & ([EEG.event.accuracy] == 0) & ([EEG.event.responded] == 1) & ([EEG.event.validRt] == 1) & ([EEG.event.extraResponse] == 0) ));
+ numTrials(4) = length(find( (strcmp({EEG.event.observation}, "ns")) & (strcmp({EEG.event.eventType}, "resp")) & (strcmp({EEG.event.congruency}, "i")) & ([EEG.event.accuracy] == 1) & ([EEG.event.responded] == 1) & ([EEG.event.validRt] == 1) & ([EEG.event.extraResponse] == 0) ));
%logical test if the number of trials for each condition (numTrials vector)
%are NOTE all >= minTrials. If statement is true, then participant/file
%is skipped and for loop over files continues to next file
- % if ~(sum(numTrials >= minTrials) == length(numTrials))
- % continue
- % end
+ if ~(sum(numTrials >= minTrials) == length(numTrials))
+ continue
+ end
% loop through conditions of interest for this file (combo of event types)
%
@@ -251,6 +242,7 @@
accuracy = 0;
responded = 1;
validRt = 1;
+ extraResponse = 0;
elseif (c==2) % social correct
observation = 's';
eventType = 'resp';
@@ -258,6 +250,7 @@
accuracy = 1;
responded = 1;
validRt = 1;
+ extraResponse = 0;
elseif (c==3) % nonsocial error
observation = 'ns';
eventType = 'resp';
@@ -265,6 +258,7 @@
accuracy = 0;
responded = 1;
validRt = 1;
+ extraResponse = 0;
elseif (c==4) % nonsocial correct
observation = 'ns';
eventType = 'resp';
@@ -272,11 +266,12 @@
accuracy = 1;
responded = 1;
validRt = 1;
+ extraResponse = 0;
end
%select combintion of event types of interest based on vars above
- EEG1 = pop_selectevent( EEG, 'latency','-1<=1','observation',observation,'eventType',eventType,'congruency',congruency,'accuracy',accuracy,'responded',responded,'validRt',validRt,'deleteevents','on','deleteepochs','on','invertepochs','off');
- EEG1 = eeg_checkset( EEG1 );
+ EEG1 = pop_selectevent( EEG, 'latency', '-1<=1', 'observation', observation, 'eventType', eventType, 'congruency', congruency, 'accuracy', accuracy, 'responded', responded, 'validRt', validRt, 'extraResponse', extraResponse, 'deleteevents', 'on', 'deleteepochs', 'on', 'invertepochs', 'off');
+ EEG1 = eeg_checkset(EEG1);
% Average across epoch dimension
% this all Channel ERP only needs to be computed once
@@ -300,4 +295,4 @@
end
%save the erps and subject list
-save('thrive_Resp_erps_min_8t_60acc.mat','erpDat_data', 'erpDat_subIds')
\ No newline at end of file
+save(sprintf('thrive_Resp_erps_min_6t_60acc_%s.mat', datestr(now, 'mm_dd_yyyy_HH_MM_SS')), 'erpDat_data', 'erpDat_subIds')
diff --git a/code/preprocessing-eeg/create_mat_csd.m b/code/preprocessing-eeg/create_mat_csd.m
index c82eab6..b4f9c88 100644
--- a/code/preprocessing-eeg/create_mat_csd.m
+++ b/code/preprocessing-eeg/create_mat_csd.m
@@ -31,16 +31,16 @@
analysis_dir = '/home/data/NDClab/analyses/thrive-theta-ddm';
%location of dataset folder
-% dataset_dir = '/Users/fzaki001/thrive-theta-ddm';
-dataset_dir = '/home/data/NDClab/datasets/thrive-dataset';
+dataset_dir = '/home/data/NDClab/analyses/thrive-theta-ddm';
+%dataset_dir = '/home/data/NDClab/datasets/thrive-dataset';
% summary_csv_path = '/Users/fzaki001/thrive-theta-ddm/derivatives/behavior/summary.csv';
summary_csv_path = '/home/data/NDClab/analyses/thrive-theta-ddm/derivatives/behavior/summary.csv';
% Setting up other things
% 1. Enter the path of the folder that has the data to be analyzed
-data_location = [dataset_dir filesep 'derivatives' filesep 'preprocessed'];
-
+%data_location = [dataset_dir filesep 'derivatives' filesep 'preprocessed'];
+data_location = [dataset_dir filesep 'derivatives' filesep 'preprocessed' filesep 'csd_data'];
% 2. Enter the path of the folder where you want to save the postprocessing outputs
output_location = [analysis_dir filesep 'derivatives' filesep 'preprocessed/erp_check'];
@@ -58,7 +58,8 @@
visitFileName = 's1_r1_e1'; %file names include "e1" designation
% Read files to analyses
-datafile_info=dir([data_location filesep 'sub-*' filesep visitDirName filesep 'eeg' filesep 'sub-*_' task '_eeg_*' procStage '_' visitFileName '.set']);
+%datafile_info=dir([data_location filesep 'sub-*' filesep visitDirName filesep 'eeg' filesep 'sub-*_' task '_eeg_*' procStage '_' visitFileName '.set']);
+datafile_info=dir([data_location filesep 'sub-*_' task '_eeg_*' procStage '_' visitFileName '.set']);
datafile_info=datafile_info(~ismember({datafile_info.name},{'.', '..', '.DS_Store'}));
datafile_names={datafile_info.name};
datafile_paths={datafile_info.folder};
@@ -72,15 +73,15 @@
% Create output folders to save data
if exist(output_location, 'dir') == 0
- mkdir(output_location)
+ mkdir(output_location);
end
-for site = 1:64
- trodes{site} = num2str(site)
-end
-Montage_64=ExtractMontage('/home/data/NDClab/analyses/thrive-theta-ddm/code/preprocessing-eeg/64ch_bv_montage_csd.csd', trodes');
+%for site = 1:64
+% trodes{site} = num2str(site);
+%end
+%Montage_64=ExtractMontage('/home/data/NDClab/analyses/thrive-theta-ddm/code/preprocessing-eeg/64ch_bv_montage_csd.csd', trodes');
% MapMontage(Montage_64);
-[G, H] = GetGH(Montage_64);
+%[G, H] = GetGH(Montage_64);
%% Count trials
% switch to output directory
@@ -90,7 +91,7 @@
% outputHeader = {'id, s_resp_incon_error, s_resp_incon_corr, ns_resp_incon_error, ns_resp_incon_corr'};
% dlmwrite(strcat('thrive_trialCounts_respOnly', date, '.csv'), outputHeader, 'delimiter', '', '-append');
-diary(sprintf('erp_log_%s.log', datestr(now, 'mm-dd-yyyy_HH_MM_SS')))
+diary(sprintf('erp_log_%s.log', datestr(now, 'mm_dd_yyyy_HH_MM_SS')))
%% pull resp-locked erp mat file
@@ -125,16 +126,6 @@
%find row in behavior file corresponding to this participant
behavior_id_match_idxs = find(behavior_info{:,'sub'} == str2num(subNumText));
- %if participant has low accuracy in either condition, skip that
- %participant for ALL conditions
- % if (behavior_info{behavior_id_match_idxs,'acc_nonsoc'} < acc_cutoff || behavior_info{behavior_id_match_idxs,'acc_soc'} < acc_cutoff)
- % continue
- % end
- %
- % if (behavior_info{behavior_id_match_idxs,'x6_or_more_err_nonsoc'} < 6 || behavior_info{behavior_id_match_idxs,'x6_or_more_err_soc'} < 6)
- % continue
- % end
- %load the original data set
EEG = pop_loadset( 'filename', datafile_names{subject}, 'filepath', datafile_paths{subject});
EEG = eeg_checkset( EEG );
@@ -142,34 +133,20 @@
EEG = pop_selectevent( EEG, 'latency','-.1 <= .1','deleteevents','on');
EEG = eeg_checkset( EEG );
- for ne = 1:length(EEG.epoch)
- myEEG = single(EEG.data(:, :, ne));
- MyResults = CSD(myEEG, G, H); % compute CSD for <channels-by-samples> 2-D epoch
- data(:, :, ne) = MyResults;
- end
- EEG.data = data
-
- data(:,:,:) = NaN;
-
- % NOTE %
- %the logic of checking conditions and then looping over conditions
- %below is fairly hard-coded and could be be substantially improved to
- %allow for easier reuse when number of conditions or number of
- %variables per condition changes.
- %
- %before pulling trials of interest, for any conditions, check to make
- %sure this file/participant has more than minTrials for EACH condition.
- %If the file/participant is below minTrials for even one of the
- %conditions that will be pulled, then the file/participant is skipped
- %entirely and no condition data at all will be pulled for this specific
- %file (but the participant can still have data pulled for another one
- %of their files from another visit).
- %
+ % for ne = 1:length(EEG.epoch)
+ % myEEG = single(EEG.data(:, :, ne));
+ % MyResults = CSD(myEEG, G, H); % compute CSD for <channels-by-samples> 2-D epoch
+ % data(:, :, ne) = MyResults;
+ % end
+ % EEG.data = data
+
+ % data(:,:,:) = NaN;
+
%count trials for each condition of interest and store in numTrials vector
- numTrials(1) = length(find( (strcmp({EEG.event.observation}, "s")) & (strcmp({EEG.event.eventType}, "resp")) & (strcmp({EEG.event.congruency}, "i")) & ([EEG.event.accuracy] == 0) & ([EEG.event.responded] == 1) & ([EEG.event.validRt] == 1) ));
- numTrials(2) = length(find( (strcmp({EEG.event.observation}, "s")) & (strcmp({EEG.event.eventType}, "resp")) & (strcmp({EEG.event.congruency}, "i")) & ([EEG.event.accuracy] == 1) & ([EEG.event.responded] == 1) & ([EEG.event.validRt] == 1) ));
- numTrials(3) = length(find( (strcmp({EEG.event.observation}, "ns")) & (strcmp({EEG.event.eventType}, "resp")) & (strcmp({EEG.event.congruency}, "i")) & ([EEG.event.accuracy] == 0) & ([EEG.event.responded] == 1) & ([EEG.event.validRt] == 1) ));
- numTrials(4) = length(find( (strcmp({EEG.event.observation}, "ns")) & (strcmp({EEG.event.eventType}, "resp")) & (strcmp({EEG.event.congruency}, "i")) & ([EEG.event.accuracy] == 1) & ([EEG.event.responded] == 1) & ([EEG.event.validRt] == 1) ));
+ numTrials(1) = length(find( (strcmp({EEG.event.observation}, "s")) & (strcmp({EEG.event.eventType}, "resp")) & (strcmp({EEG.event.congruency}, "i")) & ([EEG.event.accuracy] == 0) & ([EEG.event.responded] == 1) & ([EEG.event.validRt] == 1) & ([EEG.event.extraResponse] == 0) ));
+ numTrials(2) = length(find( (strcmp({EEG.event.observation}, "s")) & (strcmp({EEG.event.eventType}, "resp")) & (strcmp({EEG.event.congruency}, "i")) & ([EEG.event.accuracy] == 1) & ([EEG.event.responded] == 1) & ([EEG.event.validRt] == 1) & ([EEG.event.extraResponse] == 0) ));
+ numTrials(3) = length(find( (strcmp({EEG.event.observation}, "ns")) & (strcmp({EEG.event.eventType}, "resp")) & (strcmp({EEG.event.congruency}, "i")) & ([EEG.event.accuracy] == 0) & ([EEG.event.responded] == 1) & ([EEG.event.validRt] == 1) & ([EEG.event.extraResponse] == 0) ));
+ numTrials(4) = length(find( (strcmp({EEG.event.observation}, "ns")) & (strcmp({EEG.event.eventType}, "resp")) & (strcmp({EEG.event.congruency}, "i")) & ([EEG.event.accuracy] == 1) & ([EEG.event.responded] == 1) & ([EEG.event.validRt] == 1) & ([EEG.event.extraResponse] == 0) ));
%logical test if the number of trials for each condition (numTrials vector)
%are NOTE all >= minTrials. If statement is true, then participant/file
@@ -178,13 +155,7 @@
continue
end
- % loop through conditions of interest for this file (combo of event types)
- %
- % specify number of conditions using a seperate conditionNums var, so
- % that it can be referenced below when iterating idx counters (to only
- %iterate when c == length(conditionNums);
conditionNums = 1:4;
- %
for c = conditionNums
if (c==1) % social error
@@ -194,6 +165,7 @@
accuracy = 0;
responded = 1;
validRt = 1;
+ extraResponse = 0;
elseif (c==2) % social correct
observation = 's';
eventType = 'resp';
@@ -201,6 +173,7 @@
accuracy = 1;
responded = 1;
validRt = 1;
+ extraResponse = 0;
elseif (c==3) % nonsocial error
observation = 'ns';
eventType = 'resp';
@@ -208,6 +181,7 @@
accuracy = 0;
responded = 1;
validRt = 1;
+ extraResponse = 0;
elseif (c==4) % nonsocial correct
observation = 'ns';
eventType = 'resp';
@@ -215,10 +189,11 @@
accuracy = 1;
responded = 1;
validRt = 1;
+ extraResponse = 0;
end
- %select combintion of event types of interest based on vars above
- EEG1 = pop_selectevent( EEG, 'latency','-1<=1','observation',observation,'eventType',eventType,'congruency',congruency,'accuracy',accuracy,'responded',responded,'validRt',validRt,'deleteevents','on','deleteepochs','on','invertepochs','off');
+ %select combination of event types of interest based on vars above
+ EEG1 = pop_selectevent( EEG, 'latency','-1<=1','observation', observation, 'eventType', eventType, 'congruency', congruency, 'accuracy', accuracy, 'responded', responded,'validRt', validRt, 'extraResponse', extraResponse, 'deleteevents','on','deleteepochs','on','invertepochs','off');
EEG1 = eeg_checkset( EEG1 );
% Average across epoch dimension
@@ -243,4 +218,4 @@
end
%save the erps and subject list
-save('thrive_Resp_erps_csd_min_6t_60acc.mat','erpDat_data', 'erpDat_subIds')
+save(sprintf('thrive_Resp_erps_csd_min_6t_60acc_%s.mat', datestr(now, 'mm_dd_yyyy_HH_MM_SS')), 'erpDat_data', 'erpDat_subIds')
diff --git a/code/preprocessing-eeg/tf/MainScript_Calculate_TF_ITPS_ICPS.m b/code/preprocessing-eeg/tf/MainScript_Calculate_TF_ITPS_ICPS.m
index 5dffa4b..ffb6413 100644
--- a/code/preprocessing-eeg/tf/MainScript_Calculate_TF_ITPS_ICPS.m
+++ b/code/preprocessing-eeg/tf/MainScript_Calculate_TF_ITPS_ICPS.m
@@ -1,25 +1,30 @@
%%to Run on FIU HPC%
% create a local cluster object
-cluster = parcluster('local');
+%cluster = parcluster('local');
% start matlabpool with max workers set in the slurm file
-parpool(cluster, str2num(getenv('SLURM_CPUS_ON_NODE')))
+%parpool(cluster, str2num(getenv('SLURM_CPUS_ON_NODE')))
%Compute TF, ITPS, ICPS, and wPLI measures for EEG data
%Maureen Bowers 6/29/2021 based on scripts by Ranjan Debnath
-% Kianoosh Hosseini has edited this script for mfe_b study on 07/18/2023!
+
clear all;
clc;
%%
%%%%% Setting paths %%%%%
-main_dir = '/home/data/NDClab/datasets/thrive-theta-ddm'; %directory on the HPC
+main_dir = '/home/data/NDClab/analyses/thrive-theta-ddm'; %directory on the HPC
%1. Data Location
data_location = [main_dir filesep 'derivatives' filesep 'preprocessed' filesep 'csd_data'];
%2. Save Data Location
-save_location = [main_dir filesep 'derivatives' filesep 'preprocessed' filesep 'eeg' filesep 'TF_outputs' filesep 'main' filesep ];
+save_location = [main_dir filesep 'derivatives' filesep 'preprocessed' filesep 'TF_outputs' filesep 'test_fix_seed'];
+%disp(save_location)
+% Create output folders to save data
+if exist(save_location, 'dir') == 0
+ mkdir(save_location);
+end
%3. Scripts Location
scripts_location = [main_dir filesep 'code' filesep 'preprocessing-eeg' filesep 'tf'];
@@ -122,10 +127,15 @@
eeglab % Loading EEGLAB
+%rng('default'); % Reset random number generator
+%s = RandStream('mlfg6331_64'); % Create a random number stream
+
+rng(2,"twister")
+
%%%%%%%%%%%%%%%%%%%% COMPUTATIONS BEGIN BELOW HERE %%%%%%%%%%%%%%%%
%% loop through all subject
-for sub=1:5
-
+for sub=1:2
+ % Use the same stream for all workers
% Initialize objects for this participant:
timefreqs_data = [];
phase_data=[];
@@ -141,8 +151,8 @@
EEG=pop_loadset('filename', [subject], 'filepath', data_location);
EEG = pop_selectevent( EEG, 'latency','-.1 <= .1','deleteevents','on');
- EEG = pop_editeventfield( EEG, 'indices', strcat('1:', int2str(length(EEG.event))), 'Condition','NaN');
- EEG = eeg_checkset( EEG );
+ EEG = pop_editeventfield(EEG, 'indices', strcat('1:', int2str(length(EEG.event))), 'Condition','NaN');
+ EEG = eeg_checkset(EEG);
for t=1:length(EEG.event)
@@ -181,8 +191,8 @@
% Trials that are labeled as responded and validRt (trials that have larger than 150 ms RT)
% will only be included.
try
- EEG = pop_selectevent( EEG, 'validRt', 1, 'extraResponse', 0, deleteevents','on','deleteepochs','on','invertepochs','off');
- EEG = eeg_checkset( EEG );
+ EEG = pop_selectevent(EEG, 'validRt', 1, 'extraResponse', 0, 'deleteevents', 'on', 'deleteepochs', 'on', 'invertepochs', 'off');
+ EEG = eeg_checkset(EEG);
catch ME
warning('Error occurred when selecting only valid trials for subject %d: %s', subject, ME.message);
@@ -190,8 +200,8 @@
end
try
- EEG = pop_selectevent( EEG, 'responded',1, 'deleteevents','on','deleteepochs','on','invertepochs','off');
- EEG = eeg_checkset( EEG );
+ EEG = pop_selectevent(EEG, 'responded', 1, 'deleteevents', 'on', 'deleteepochs', 'on', 'invertepochs', 'off');
+ EEG = eeg_checkset(EEG);
catch ME
warning('Error occurred when selecting only valid trials (responded) for subject %d: %s', subject, ME.message);
@@ -214,4 +224,4 @@
%Save out table with trial numbers
TrialNums_table = struct2table(TrialNums);
-writetable(TrialNums_table,[save_location 'TrialNums.csv']);
\ No newline at end of file
+writetable(TrialNums_table,[save_location 'TrialNums.csv']);
diff --git a/code/preprocessing-eeg/tf/MainScript_Calculate_TF_ITPS_ICPS_para.m b/code/preprocessing-eeg/tf/MainScript_Calculate_TF_ITPS_ICPS_para.m
index 85dd898..e402950 100644
--- a/code/preprocessing-eeg/tf/MainScript_Calculate_TF_ITPS_ICPS_para.m
+++ b/code/preprocessing-eeg/tf/MainScript_Calculate_TF_ITPS_ICPS_para.m
@@ -19,7 +19,7 @@
data_location = [main_dir filesep 'derivatives' filesep 'preprocessed' filesep 'csd_data'];
%2. Save Data Location
-save_location = [main_dir filesep 'derivatives' filesep 'preprocessed' filesep 'TF_outputs' filesep 'main' filesep 'resp' filesep];
+save_location = [main_dir filesep 'derivatives' filesep 'preprocessed' filesep 'TF_outputs' filesep 'main' filesep 'stim' filesep];
%disp(save_location)
% Create output folders to save data
if exist(save_location, 'dir') == 0
@@ -51,20 +51,33 @@
%7. What are your conditions of interest if using Event-Related Data? This
%naming convention should come from the Edit_events.m script provided.
%Note: not needed for resting state data.
-
+%Super_Conds = {
+%{'resp_s_i_0'},...
+%{'resp_s_i_1'},...
+%{'resp_s_c_1'},...
+%{'resp_ns_i_0'},...
+%{'resp_ns_i_1'},...
+%{'resp_ns_c_1'},...
+%}
% RESPONSE CONDITIONS
-Conds = {
+%Conds = {
%'resp_s_i_0',...
%'resp_s_i_1',...
%'resp_s_c_1',...
%'resp_ns_i_0',...
%'resp_ns_i_1',...
-'resp_ns_c_1',...
-}; % resp_i_0 = incong error response; resp_i_1 = incong correct response
+%'resp_ns_c_1',...
+%}; % resp_i_0 = incong error response; resp_i_1 = incong correct response
% STIMULUS CONDITIONS
-%Conds = {'stim_s_i_0','stim_s_i_1', 'stim_s_c_1', ...
-% 'stim_ns_i_0','stim_ns_i_1', 'stim_ns_c_1',}; % resp_i_0 = incong error stim; resp_i_1 = incong correct response
+Conds = {
+%'stim_s_i_0',...
+%'stim_s_i_1',...
+%'stim_s_c_1',...
+%'stim_ns_i_0',...
+%'stim_ns_i_1',...
+'stim_ns_c_1',...
+}; % resp_i_0 = incong error stim; resp_i_1 = incong correct response
%8. Minimum number of trials to analyze
mintrialnum = 6; %If the participant does not have enough trials in a condition based on this cutoff, a "notenoughdata.mat" file will be saved into save_location.
@@ -97,7 +110,7 @@
%%%%% Questions about phase-based measures %%%%%
%14. Would you like to calculate inter-trial phase synchrony (ITPS) in addition to TF?
-ITPS_calc = 0; %(1=yes,0=no)
+ITPS_calc = 1; %(1=yes,0=no)
%15. Would you like to subsample trials? This is recommended for event-related paradigms,
% especially when there are uneven numbers of trials in conditions.
@@ -114,7 +127,7 @@
ICPS_calc = 1; %(1=yes,0=no)
%17. Would you like to calculate coherence or weighted phaselagidx?
-ICPS_or_wPLI = 0; %(1=coherence, 0=wPLI)
+ICPS_or_wPLI = 1; %(1=coherence, 0=wPLI)
%18. Inter-channel phase synchrony over trials or connectivity over time?
% Kia: over trials results in a frequency by time matrix and over time
@@ -146,12 +159,8 @@
TrialNums = zeros(length(subject_list), length(Conds));
%rng(2, 'twister'); % to fix seed
-%D = parallel.pool.DataQueue;
-%waitbar = waitbar(0, 'Processing subjects...');
-%afterEach(D, @(x) waitbar(x/length(subject_list), waitbar));
%%%%%%%%%%%%%%%%%%%% COMPUTATIONS BEGIN BELOW HERE %%%%%%%%%%%%%%%%
%% loop through all subject
-
parfor sub=1:length(subject_list)
%parfor sub=1:2
% rng(2, 'twister'); % if you want to fix the seed put it inside the loop as well
diff --git a/code/preprocessing-eeg/tf/timefreq_script.m b/code/preprocessing-eeg/tf/timefreq_script.m
index 249d36a..601f1b0 100644
--- a/code/preprocessing-eeg/tf/timefreq_script.m
+++ b/code/preprocessing-eeg/tf/timefreq_script.m
@@ -190,8 +190,9 @@
%pull out condition data
cond_data = []; cond_EEG=[];
- cond_EEG = pop_selectevent( EEG, 'Condition',(Conds{cond}), 'deleteevents','on','deleteepochs','on','invertepochs','off');
- cond_data = eeg_checkset( cond_EEG.data );%add pulling out condition specific data
+ cond_EEG = pop_selectevent(EEG, 'Condition',(Conds{cond}), 'deleteevents','on','deleteepochs','on','invertepochs','off');
+ cond_data = eeg_checkset(cond_EEG.data);%add pulling out condition specific data
+ fprintf('\n* Condition %s data shape: %d\n *\n', Conds{cond}, size(cond_data, 3));
%Create TrialNums structure to be saved out
TrialNums( sub+(cond-1)+((sub-1)*(length(Conds)-1)) ).subject = subject;
diff --git a/code/preprocessing-eeg/transform_csd.m b/code/preprocessing-eeg/transform_csd.m
index 648643a..c040c3f 100644
--- a/code/preprocessing-eeg/transform_csd.m
+++ b/code/preprocessing-eeg/transform_csd.m
@@ -5,11 +5,17 @@
% Workshop on 02/22. This script uses parts of the "set up" structure from
% the MADE preprocessing pipeline (Debnath, Buzzell, et. al., 2020)
-clear % clear matlab workspace
-clc % clear matlab command window
-
+%clear % clear matlab workspace
+%clc % clear matlab command window
+cluster = parcluster('local');
%% Setting up other things
-
+parpool(cluster, str2num(getenv('SLURM_CPUS_PER_TASK'))) % this should be same as --cpus-per-task
+pool = gcp('nocreate'); % Get the current parallel pool without creating a new one
+if isempty(pool)
+ disp('No parallel pool is currently running.');
+else
+ disp(['Parallel pool with ', num2str(pool.NumWorkers), ' workers is running.']);
+end
%Location of MADE and ADJUSTED-ADJUST scripts
% addpath(genpath([main_dir filesep 'MADE-EEG-preprocessing-pipeline']));% enter the path of the EEGLAB folder in this line
addpath(genpath('/home/data/NDClab/tools/lab-devOps/scripts/MADE_pipeline_standard/eeg_preprocessing'));% enter the path of the folder in this line
@@ -71,35 +77,41 @@
trodes{site} = num2str(site);
end
Montage_64=ExtractMontage('/home/data/NDClab/analyses/thrive-theta-ddm/code/preprocessing-eeg/64ch_bv_montage_csd.csd', trodes');
-% MapMontage(Montage_64);
[G, H] = GetGH(Montage_64);
diary(sprintf('csd_log_%s.log', datestr(now, 'mm-dd-yyyy_HH_MM_SS')))
% loop through each participant in the study
-for subject = 1:length(datafile_names)
-
- % extract participant number
- subNumText = datafile_names{subject}(5:11);
-
- %load the original data set
- EEG = pop_loadset('filename', datafile_names{subject}, 'filepath', datafile_paths{subject});
- EEG = eeg_checkset(EEG);
- disp(datafile_names{subject})
- disp(save_location)
- %remove all the non-stim-locking markers (should have done already...)
- EEG = pop_selectevent(EEG, 'latency','-.1 <= .1','deleteevents','on');
- EEG = eeg_checkset(EEG);
-
- for ne = 1:length(EEG.epoch)
- myEEG = single(EEG.data(:, :, ne));
- MyResults = CSD(myEEG, G, H); % compute CSD for <channels-by-samples> 2-D epoch
- data(:, :, ne) = MyResults;
+parfor subject = 1:length(datafile_names)
+ try
+ % extract participant number
+ subNumText = datafile_names{subject}(5:11);
+
+ %load the original data set
+ EEG = pop_loadset('filename', datafile_names{subject}, 'filepath', datafile_paths{subject});
+ EEG = eeg_checkset(EEG);
+
+ %remove all the non-stim-locking markers (should have done already...)
+ EEG = pop_selectevent(EEG, 'latency','-.1 <= .1','deleteevents','on');
+ EEG = eeg_checkset(EEG);
+ fprintf('Subject %s: Processing %d events\n', subNumText, length(EEG.event));
+ data = zeros(size(EEG.data), 'single');
+ for ne = 1:length(EEG.epoch)
+ myEEG = single(EEG.data(:, :, ne));
+ MyResults = CSD(myEEG, G, H); % compute CSD for <channels-by-samples> 2-D epoch
+ data(:, :, ne) = MyResults;
+ end
+
+ EEG.data = data;
+ EEG = eeg_checkset(EEG);
+ disp(size(EEG.event))
+
+ EEG = pop_editset(EEG, 'setname', datafile_names{subject});
+ EEG = pop_saveset(EEG, 'filename', datafile_names{subject}, 'filepath', save_location);
+ fprintf('Subject %s: Successfully processed and saved\n', subNumText);
+% clear data;
+ catch ME
+ fprintf('Error processing subject %s: %s\n', subNumText, ME.message);
+ continue;
end
- EEG.data = data;
-
- data(:,:,:) = NaN;
-
- EEG = pop_editset(EEG, 'setname', datafile_names{subject});
- EEG = pop_saveset(EEG, 'filename', datafile_names{subject}, 'filepath', save_location);
end