78 lines
1.7 KiB
Matlab
78 lines
1.7 KiB
Matlab
% test script
|
|
|
|
c3 = 14;
|
|
c4 = 18;
|
|
|
|
mybasis = create_bspline_basis([0 .5], 30, 4);
|
|
|
|
load ~/brains/bci_comp_2003_dataSet_IV/sp1s_aa_1000Hz;
|
|
|
|
|
|
|
|
x_train = CAR(x_train);
|
|
ground_truth = y_train;
|
|
|
|
num_train = 216;
|
|
% num_train = 316;
|
|
|
|
y_train = squeeze(x_train(:,c3,1:num_train));
|
|
y_test = squeeze(x_train(:,c3,217:end));
|
|
|
|
y_train_mean = mean(y_train,2);
|
|
|
|
y_train = y_train - repmat(y_train_mean, [1 num_train]);
|
|
y_test = y_test - repmat(y_train_mean, [1 100]);
|
|
|
|
% y_shifted = x_train_c3_shifted;
|
|
|
|
% y is num samples (500) by num epochs (318)
|
|
argvals = .001:.001:.5;
|
|
|
|
myfd_train = data2fd(y_train, argvals, mybasis);
|
|
myfd_test = data2fd(y_test, argvals, mybasis);
|
|
%myfd_shifted = data2fd(y_shifted, argvals, mybasis);
|
|
|
|
%pca_results = pca_fd(myfd_train, 30);
|
|
pca_results = pca_fd(myfd_train, 30, fdPar(mybasis, 2, 0));
|
|
|
|
% treat left and right data differently
|
|
|
|
|
|
harmfd = struct(pca_results.harmfd);
|
|
score_train = pca_results.harmscr;
|
|
smyfd_train = struct(myfd_train);
|
|
smyfd_test = struct(myfd_test);
|
|
|
|
|
|
for i=1:100
|
|
score_test(i,:) = smyfd_test.coef(:,i)' * inv(harmfd.coef');
|
|
end
|
|
|
|
|
|
|
|
|
|
|
|
lefts_train = find(ground_truth(1:num_train) == 0);
|
|
rights_train = find(ground_truth(1:num_train) == 1);
|
|
|
|
lefts_test = find(ground_truth((num_train+1):end) == 0);
|
|
rights_test = find(ground_truth((num_train+1):end) == 1);
|
|
|
|
|
|
pc = 2;
|
|
|
|
lmean = mean(score_train(lefts_train, pc));
|
|
rmean = mean(score_train(rights_train, pc));
|
|
|
|
|
|
left_correct = ...
|
|
abs(score_test(lefts_test, pc) - lmean) < ...
|
|
abs(score_test(lefts_test, pc) - rmean);
|
|
|
|
right_correct = ...
|
|
abs(score_test(rights_test, pc) - rmean) < ...
|
|
abs(score_test(rights_test, pc) - lmean);
|
|
|
|
[sum(left_correct) / length(left_correct)
|
|
sum(right_correct) / length(right_correct)]
|