clc
clear all
close all
%%
I1 = importdata('WEAKV1.txt');
V = I1(:,2) ;
time = I1(:,1) ;
%%
load ('V')
load ('N')
%%
% I1 = importdata('V22.txt');
% V = I1(:,2) ;
% time = I1(:,1) ;
%
I2 = importdata('N22.txt');
N = I2(:,2) ;
W=zeros(length(N),1) ;
MEAN = mean(N) ;
N = N - MEAN ;
%%
ms = 10 ;
T = 160*10 ;
c=1 ;
p=1 ;
w=1 ;
% figure
% findpeaks(V,'MinPeakDistance',1600, 'MinPeakHeight',20) ;
% hold on
%%
[pks , locs] = findpeaks(V,'MinPeakDistance',1600, 'MinPeakHeight',20) ;
%
K = find( pks>20) ;
length(K)
%
for i=3:length(pks)
if pks(i,1) > 20
A(1,:) = V(locs(i,1)-100:locs(i,1) , 1) ;
AA = A(A>=-20) ;
[L]=knnsearch(AA',-20) ;
B = AA(1,L) ;
[row1]=find( V==B) ;
for jjj=1:size(row1,1)
if (row1(jjj,1) < locs(i,1)-5) && (row1(jjj,1) > locs(i,1)-100)
row2 = row1(jjj,1) ;
S(:,c)= N(row2-T:row2,1) ;
c=c+1 ;
W(row2,1) = 1 ;
% plot(row2, V(row2) , '*g')
% hold on
end
end
end
end
save ('S.mat' , 'S')
save ('W.mat' , 'W')
[m,n]=size(S) ;
NumberSpike = n
%%
stc_Matrix = S ;
m_s = mean(stc_Matrix,2);
time = -time(1:160*10+1,1) ;
time = sort(time) ;
%%
cc=1 ;
spp = W ;
sp_binary_shuf(:,1) = Shuffle_(spp);
spike_indx = find( sp_binary_shuf >0);
maxindx = spike_indx(end);
indx_int = find(spike_indx>T & spike_indx<maxindx-T);
for i=1:length( indx_int )
SS(:,cc)= N( spike_indx(indx_int(i)) -T: spike_indx(indx_int(i)) ,1) ;
cc=cc+1 ;
end
stc_Matrix_shuf = SS ;
%%
ccc=1 ;
spp = W ;
sp_binary_shuf2(:,1) = Shuffle_(spp);
spike_indx2 = find( sp_binary_shuf2 >0);
maxindx2 = spike_indx2(end);
indx_int2 = find(spike_indx2>T & spike_indx2<maxindx2-T);
for i=1:length( indx_int2 )
SSS(:,ccc)= N( spike_indx2(indx_int2(i)) -T: spike_indx2(indx_int2(i)) ,1) ;
ccc=ccc+1 ;
end
stc_Matrix_shuf2 = SSS ;
%%
inp_mean = mean(stc_Matrix,2); INP_mean = repmat(inp_mean,1,size(stc_Matrix,2));
inp_mean_shuf = mean(stc_Matrix_shuf,2); INP_mean_shuf = repmat(inp_mean_shuf,1,size(stc_Matrix_shuf,2));
inp_mean_shuf2 = mean(stc_Matrix_shuf2,2); INP_mean_shuf2 = repmat(inp_mean_shuf2,1,size(stc_Matrix_shuf2,2));
C_shuf2 = cov((stc_Matrix_shuf2 - INP_mean_shuf2)');
C_shuf = cov((stc_Matrix_shuf - INP_mean_shuf)') - C_shuf2;
C = cov((stc_Matrix - INP_mean)') - C_shuf2;
[U_shuf,D_shuf] = eig(C_shuf);%[u_shuf,s_shuf,v] = svd(stc_new_shuf);
[U,D] = eig(C);%-C_shuf);
d = diag(D)/norm(diag(D));
v1 = U(:,1601);%
v2 = U(:,1600);%
%v3 = U(:,2);%
%%
figure
plot(time,m_s/norm(m_s),'k','LineWidth',2)
xlabel(' t before spike (ms)')
legend ('STA')
%%
figure
plot(time,m_s/norm(m_s),'k','LineWidth',2)
hold on, plot(time,v1/norm(v1),'c','LineWidth',0.8)
hold on, plot(time,v2/norm(v2),'r','LineWidth',0.8)
legend('STA','feature 1', 'feature 2')
xlabel(' t before spike (ms)')
%%
d_shuf = diag(D_shuf)/norm(diag(D_shuf));
figure
plot(length(d_shuf):-1:1,d_shuf(1:end)/norm(d_shuf,2),'ok','LineWidth',1)
hold on
plot(length(d):-1:1,d(1:end)/norm(d,2),'og','LineWidth',1)
hold on
plot(1 , d(1601)/norm(d,2),'oc','LineWidth',1)
plot(2 , d(1600)/norm(d,2),'or','LineWidth',1)
xlabel('eigenmode index')
ylabel('eigenvalue')
%%
%[s1,s2] = proj1(v1 , v2, S+0.06 ,NumberSpike );
[s1,s2] = Projection_Vector(W, v1 , v2 , N );
[s1_shuf,s2_shuf] = Projection_Vector(sp_binary_shuf,v1,v2, N);%
[s1_shuf2,s2_shuf2] = Projection_Vector(sp_binary_shuf2,v1,v2, N );%
figure
plot(s1_shuf,s2_shuf,'Color',[0.75 0.75 0.75], 'linestyle','none' , 'Marker', '.')
hold on
plot(s1_shuf2,s2_shuf2,'Color',[0.75 0.75 0.75], 'linestyle','none' , 'Marker', '.')
xlim([-0.25 0.25])
ylim([-0.25 0.25])
hold on
plot(s1,s2,'.k')
xlabel('Projection onto feature 1')
ylabel('Projection onto feature 2')
% figure
% plot(diag(D_shuf),'og')
% hold on
% plot(diag(D),'or')
% %%
%
% figure
%
% plot(time,U(:,1598)/norm(U(:,1599)),'b','LineWidth',2)
% ssp_binary_shuf = find(sp_binary_shuf2>0) ;
% [ss1 ss2]=Projection_ (ssp_binary_shuf, C_shuf2 , N+0.06) ;
%
% WW=find(W>0) ;
% [s1,s2] = Projection_(WW,C,N+0.06) ;
% figure
%
%
% plot(ss1,ss2,'.k')
% hold on
% plot(s1,s2,'.r')
%%
% S(:,5)'
% %%
% figure
% plot(v1)
% %%
% figure
% plot(S(:,50))