-
Notifications
You must be signed in to change notification settings - Fork 1
/
ASDPupil_swlm.m
202 lines (170 loc) · 7.98 KB
/
ASDPupil_swlm.m
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
%% this script uses a sliding window, and a linear model, for each subject
clearvars
close all
%% load data
root = 'D:\Ruonan\Projects in the lab\Ambiguity-as-stressor Project\Tobii script\AS_PatternPilotData\AS_DecisionTobiiData';
subj = [21 22 23 24 25 26 27 28 29 30 31 32 33 34];
datafold = fullfile(root,'Matlab data','pupildata_normalized\');
figfold = fullfile(root, 'Matlab data', 'pupil regression fig','Left right eye average_normalized');
% get screen size for the convenience of plotting
screensize = get(groot, 'Screensize');
for subjidx = 1:length(subj)
%% load data and exclude bad trials
% subjidx = 1; % for testing, should be commented out
dataname = ['ASD' num2str(subj(subjidx)) '_Initial_norm.mat'];
load([datafold,dataname])
% Choose the signal to look at
pupil = sInitial.filtered.Pupil_norm;
vel = sInitial.filtered.vel;
% data qualitiy: proportion of missing data for each trial
% first column-left, second column-right, third column-averaged
signalQual = sInitial.DataQual;
% check the data quality: after preproc, how many trials remain
% standard for excluding a trial: in the chosen time window, if over 18% data is missing
threshold = 0.3;
trial2analyze = sum(signalQual(:,3) < threshold);
pupil(signalQual(:,3) > threshold,:) = nan;
vel(signalQual(:,3) > threshold,:) = nan;
%% Linear model pupil ~ al + val, with a sliding window
x = [sInitial.AL sInitial.Val];
windl = 0; % actual length = (windl+1+windl) * 1000/60 ms
timestamp = sInitial.Timestamp(1,:);
% regression coefficient
regcoeff = zeros(length(pupil),2);
% p-value
pval = zeros(length(pupil),2);
for i = windl+1 : length(pupil)-windl
y = nanmean(pupil(:,i-windl:i+windl),2); % average of the time window
mdl = LinearModel.fit(x,y);
coeff = table2array(mdl.Coefficients);
regcoeff(i,1) = coeff(2,1); % regression coefficient for AL
regcoeff(i,2) = coeff(3,1); % regression coefficient for Val
pval(i,1) = coeff(2,4); % p value for AL
pval(i,2) = coeff(3,4); % p value for Val
end
% plot regression
valsignif = find(pval(:,2)<0.05 & pval (:,2)>0);
alsignif = find(pval(:,1)<0.05 & pval (:,1)>0);
figRegress = figure('Position', [screensize(3)/4 screensize(4)/4 screensize(3)/2 screensize(4)/2]);
% plot regression coefficient
% ambiguity level, blue
plot([1:length(regcoeff)]*1000/60,regcoeff(:,1),'Color','b')
hold on
% value, red
plot([1:length(regcoeff)]*1000/60,regcoeff(:,2),'Color','r')
yLimit = ylim;
xLimit = xlim;
% paint the windows with significance (uncorrected)
% ambiguity level, blue
if isempty(alsignif)==0
color = [0 0 1];
drawy = linspace(yLimit(1),yLimit(2));
for k = 1:length(alsignif)
drawx = repmat(alsignif(k)*1000/60,length(drawy),2);
plot(drawx,drawy,'Color',[0 0 1 0.2])
end
% aal=area(alsignif,repmat(yaxis(1),length(alsignif),1),'basevalue',yaxis(2),...
% 'FaceColor',color,'EdgeColor','none','ShowBaseLine','off');
% % make area color transparent
% drawnow; pause(0.05); % This needs to be done for transparency to work
% aal.Face.ColorType = 'truecoloralpha';
% aal.Face.ColorData(4) = 255 * 0.3; % Your alpha value is the 0.3
else
disp('No siginificant time window for al')
end
% value, red
if isempty(valsignif) ==0
color = [1 0 0];
drawy = linspace(yLimit(1),yLimit(2));
for k = 1:length(valsignif)
drawx = repmat(valsignif(k)*1000/60,length(drawy),2);
plot(drawx,drawy,'Color',[1 0 0 0.2])
end
% aval=area(valsignif,repmat(yaxis(1),length(valsignif),1),'basevalue',yaxis(2),...
% 'FaceColor',color,'EdgeColor','none','ShowBaseLine','off');
% % make area color transparent
% drawnow; pause(0.05); % This needs to be done for transparency to work
% aval.Face.ColorType = 'truecoloralpha';
% aval.Face.ColorData(4) = 255 * 0.3; % Your alpha value is the 0.3
else
disp('No siginificant time window for val')
end
txtPar1 = ['window length = ' num2str(windl*2+1) '; included trials = ' num2str(trial2analyze)];
text(xLimit(2)*5/9,yLimit(1)+(yLimit(2)-yLimit(1))/8,txtPar1)
title(['ASD ' num2str(subj(subjidx)), ' regression Pupil ~ val(red) +al(blue)'])
figname = fullfile(figfold,['ASD' num2str(subj(subjidx)), '_regression_pupilsize.fig' ]);
saveas (figRegress, figname);
bmpname = fullfile(figfold,['ASD' num2str(subj(subjidx)), '_regression_pupilsize.bmp' ]);
saveas (figRegress, bmpname);
%% Linear model pupil_velocity ~ al + val, with a sliding window
x = [sInitial.AL sInitial.Val];
windl = 0; % actual length = (windl+1+windl) * 1000/60 ms
timestamp = sInitial.Timestamp(1,:);
% pre-allocate regression coefficient
regcoeff = zeros(length(vel),2);
% pre-allocate p-value
pval = zeros(length(vel),2);
for i = windl+1 : length(vel)-windl
y = nanmean(vel(:,i-windl:i+windl),2); % average of the time window
mdl = LinearModel.fit(x,y);
coeff = table2array(mdl.Coefficients);
regcoeff(i,1) = coeff(2,1); % regression coefficient for AL
regcoeff(i,2) = coeff(3,1); % regression coefficient for Val
pval(i,1) = coeff(2,4); % p value for AL
pval(i,2) = coeff(3,4); % p value for Val
end
% plot regression
valsignif = find(pval(:,2)<0.05 & pval (:,2)>0);
alsignif = find(pval(:,1)<0.05 & pval (:,1)>0);
figRegress = figure('Position', [screensize(3)/4 screensize(4)/4 screensize(3)/2 screensize(4)/2]);
% plot regression coefficient
% ambiguity level, blue
plot([1:length(regcoeff)]*1000/60,regcoeff(:,1),'Color','b')
hold on
% value, red
plot([1:length(regcoeff)]*1000/60,regcoeff(:,2),'Color','r')
yLimit = ylim;
xLimit = xlim;
% paint the windows with significance (uncorrected)
% ambiguity level, blue
if isempty(alsignif)==0
color = [0 0 1];
drawy = linspace(yLimit(1),yLimit(2));
for k = 1:length(alsignif)
drawx = repmat(alsignif(k)*1000/60,length(drawy),2);
plot(drawx,drawy,'Color',[0 0 1 0.2])
end
% aal=area(alsignif,repmat(yaxis(1),length(alsignif),1),'basevalue',yaxis(2),...
% 'FaceColor',color,'EdgeColor','none','ShowBaseLine','off');
% % make area color transparent
% drawnow; pause(0.05); % This needs to be done for transparency to work
% aal.Face.ColorType = 'truecoloralpha';
% aal.Face.ColorData(4) = 255 * 0.3; % Your alpha value is the 0.3
else
disp('No siginificant time window for al')
end
% value, red
if isempty(valsignif) ==0
color = [1 0 0];
drawy = linspace(yLimit(1),yLimit(2));
for k = 1:length(valsignif)
drawx = repmat(valsignif(k)*1000/60,length(drawy),2);
plot(drawx,drawy,'Color',[1 0 0 0.2])
end
% aval=area(valsignif,repmat(yaxis(1),length(valsignif),1),'basevalue',yaxis(2),...
% 'FaceColor',color,'EdgeColor','none','ShowBaseLine','off');
% % make area color transparent
% drawnow; pause(0.05); % This needs to be done for transparency to work
% aval.Face.ColorType = 'truecoloralpha';
% aval.Face.ColorData(4) = 255 * 0.3; % Your alpha value is the 0.3
else
disp('No siginificant time window for val')
end
txtPar1 = ['window length = ' num2str(windl*2+1) '; included trials = ' num2str(trial2analyze)];
text(xLimit(2)*5/9,yLimit(1)+(yLimit(2)-yLimit(1))/8,txtPar1)
title(['ASD ' num2str(subj(subjidx)), ' regression PupilVelocity ~ val(red) +al(blue)'])
figname = fullfile(figfold,['ASD' num2str(subj(subjidx)), '_regression_velocity.fig' ]);
saveas (figRegress, figname);
bmpname = fullfile(figfold,['ASD' num2str(subj(subjidx)), '_regression_velocity.bmp' ]);
saveas (figRegress, bmpname);
end