-
Notifications
You must be signed in to change notification settings - Fork 1
/
pdcca_acc_tsinghua_20220401_.m
230 lines (186 loc) · 7.99 KB
/
pdcca_acc_tsinghua_20220401_.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
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
% This code is prepared by Chi Man Wong (chiman465@gmail.com)
% Date: 16 Nov 2022
clear all;
close all;
str_dir='..\TH_MFSC\';
dataset_no=2; % 1: offline exp., 2: online exp.
data_len=1;
if dataset_no==1
num_of_subj=8; % Number of subjects
elseif dataset_no==2
num_of_subj=12; % Number of subjects
else
end
Fs=250; % sample rate
ch_used=[1:9]; % Pz, PO5, PO3, POz, PO4, PO6, O1,Oz, O2 (in SSVEP benchmark dataset)
num_of_harmonics=5; % for all cca-based methods
num_of_subbands=6; % for filter bank analysis
FB_coef0=[1:num_of_subbands].^(-1.25)+0.25; % for filter bank analysis
latencyDelay = round(0.14*Fs);% time windows for CCA training
% time-window length (min_length:delta_t:max_length)
min_length=4*data_len;
delta_t=0.2;
max_length=min_length; % [min_length:delta_t:max_length]
is_center_std=0; % 0: without , 1: with (zero mean, and unity standard deviation)
%bandpass filter
for k=1:num_of_subbands
Wp = [(8*k)/(Fs/2) 90/(Fs/2)];
Ws = [(8*k-2)/(Fs/2) 100/(Fs/2)];
[N,Wn] = cheb1ord(Wp,Ws,3,40);
[subband_signal(k).bpB,subband_signal(k).bpA] = cheby1(N,0.5,Wn);
% subband(k).bpdata=zeros(length(eeg_channels),round(max(epochTime)*srate),length(condition),length(blocknum),length(sub));
end
%notch
Fo = 50;
Q = 35;
BW = (Fo/(Fs/2))/Q;
[notchB,notchA] = iircomb(Fs/Fo,BW,'notch');
seed = RandStream('mt19937ar','Seed','shuffle');
mycounter=1;
load([str_dir 'reqCodeword.mat']);
for sn=1:num_of_subj
tic
if dataset_no==1
load(strcat(str_dir,'\Offline\S',num2str(sn),'.mat'));
elseif dataset_no==2
load(strcat(str_dir,'\Online\S',num2str(sn),'.mat'));
else
end
% pre-stimulus period: 0.5 sec
% latency period: 0.14 sec
eeg=data(ch_used,:,:,:);
[d1_,d2_,d3_,d4_]=size(eeg);
d1=d3_;d2=d4_;d3=d1_;d4=d2_;
no_of_class=d1;
% d1: num of stimuli
% d2: num of trials
% d3: num of channels % Pz, PO5, PO3, POz, PO4, PO6, O1, Oz, O2
% d4: num of sampling points
for i=1:1:d1
for j=1:1:d2
y0=reshape(eeg(:,:,i,j),d3,d4);
SSVEPdata(:,:,j,i)=reshape(y0,d3,d4,1,1);
y = filtfilt(notchB, notchA, y0.'); %notch
y = y.';
for sub_band=1:num_of_subbands
for ch_no=1:d3
tmp2=filtfilt(subband_signal(sub_band).bpB,subband_signal(sub_band).bpA,y(ch_no,:));
y_sb(ch_no,:) = tmp2(latencyDelay+1:latencyDelay+4*Fs);
end
subband_signal(sub_band).SSVEPdata(:,:,j,i)=reshape(y_sb,d3,length(y_sb),1,1);
end
end
end
clear eeg
%% Initialization
n_ch=size(SSVEPdata,1);
TW=min_length:delta_t:max_length;
TW_p=round(TW*Fs);
n_run=d2; % number of used runs
pha_val=[0 0.5 1 1.5 0 0.5 1 1.5]*pi;
sti_f=[8:15];
n_sti=length(sti_f); % number of stimulus frequencies
for tw_length=1:length(TW)
sig_len=floor(TW_p(tw_length)/4);
for j=1:no_of_class
ref4=[];ref4a=[];
for k=1:4
jj=reqCodeword(j,k)+1;
ref0=ref_signal_nh(sti_f(jj),Fs,pha_val(jj),sig_len,num_of_harmonics);
ref4=[ref4 ref0];
ref0=ref_signal_nh(sti_f(jj),Fs,0,sig_len,num_of_harmonics);
ref4a=[ref4a ref0];
end
NewYref{tw_length}(:,:,j)=ref4;
NewYrefa{tw_length}(:,:,j)=ref4a;
end
for j=1:length(sti_f)
ref1=ref_signal_nh(sti_f(j),Fs,pha_val(j),sig_len,num_of_harmonics);
Yref{tw_length}(:,:,j)=ref1;
end
end
FB_coef=FB_coef0'*ones(1,no_of_class);
n_correct=zeros(length(TW),5); % Count how many correct detection
idx_testdata=1:n_run; % index of the testing trials
for run_test=1:length(idx_testdata)
for tw_length=1:length(TW)
sig_len=TW_p(tw_length);
seg_len=floor(TW_p(tw_length)/4);
% test_signal=zeros(d3,sig_len);
fprintf('Testing TW %fs, Run No. %d , Dataset No. %d \n',TW(tw_length),idx_testdata(run_test),dataset_no);
old_test_signal=zeros(d3,sig_len,num_of_subbands);
NACCAR=[];
for i=1:no_of_class
for sub_band=1:num_of_subbands
test_signal=subband_signal(sub_band).SSVEPdata(:,1:4*Fs,idx_testdata(run_test),i);
if (is_center_std==1)
test_signal=test_signal-mean(test_signal,2)*ones(1,length(test_signal));
test_signal=test_signal./(std(test_signal')'*ones(1,length(test_signal)));
end
for j=1:no_of_class
Y=NewYref{tw_length}(:,:,j);
[A,B,r]=canoncorr(test_signal(:,[1:seg_len,Fs+1:Fs+seg_len,2*Fs+1:2*Fs+seg_len,3*Fs+1:3*Fs+seg_len])',Y');
mscca_r(sub_band,j)=r(1)*FB_coef0(sub_band);
Y=NewYrefa{tw_length}(:,:,j);
[A,B,r]=canoncorr(test_signal(:,[1:seg_len,Fs+1:Fs+seg_len,2*Fs+1:2*Fs+seg_len,3*Fs+1:3*Fs+seg_len])',Y');
mscca_r2(sub_band,j)=r(1)*FB_coef0(sub_band);
end
for k=1:4
X=test_signal(:,(k-1)*Fs+1:(k-1)*Fs+seg_len);
for j=1:n_sti
Y=Yref{tw_length}(:,:,j);
[A,B,r]=canoncorr(X',Y');
cca_r(sub_band,j,k)=r(1)*FB_coef0(sub_band);
end
end
end
MSCCAR=squeeze(sum(mscca_r,1));
[~,idx]=max(MSCCAR);
if idx==i
n_correct(tw_length,3)=n_correct(tw_length,3)+1;
end
MSCCAR2=squeeze(sum(mscca_r2,1));
[~,idx]=max(MSCCAR2);
if idx==i
n_correct(tw_length,2)=n_correct(tw_length,2)+1;
end
cca_r1=squeeze(sum(cca_r,1));
for j=1:no_of_class
CCAR(j)=sqrt(cca_r1(reqCodeword(j,1)+1,1)^2+...
cca_r1(reqCodeword(j,2)+1,2)^2+...
cca_r1(reqCodeword(j,3)+1,3)^2+...
cca_r1(reqCodeword(j,4)+1,4)^2);
end
[~,idx]=max(CCAR);
if idx==i
n_correct(tw_length,1)=n_correct(tw_length,1)+1;
end
MSCCAR3=MSCCAR+CCAR/4;
[~,idx]=max(MSCCAR3);
if idx==i
n_correct(tw_length,5)=n_correct(tw_length,5)+1;
end
MSCCAR4=MSCCAR2+CCAR/4;
[~,idx]=max(MSCCAR4);
if idx==i
n_correct(tw_length,4)=n_correct(tw_length,4)+1;
end
end
end
end
%% Save results
toc
accuracy=100*n_correct/no_of_class/length(idx_testdata)
for tw_length=1:length(TW)
itr(tw_length,:)=itr_bci(accuracy(tw_length,:)/100,no_of_class,(TW(tw_length)*ones(1,5)+0.5));
end
itr
% column 1: CCA
% column 2: pdCCA0
% column 3: pdCCA
% column 4: pdCCA0+
% column 5: pdCCA+
xlswrite('acc_file.xlsx',accuracy'/100,strcat('Sheet',num2str(sn)));
xlswrite('itr_file.xlsx',itr',strcat('Sheet',num2str(sn)));
disp(sn)
end