two_group_test_spectrum

 function [dz,vdz,Adz]=two_group_test_spectrum(J1,J2,p,plt,f)
 Test the null hypothesis (H0) that data sets J1, J2 in 
 two conditions c1,c2 have equal population spectrum

 Usage:
 [dz,vdz,Adz]=two_sample_test_spectrum(J1,J2,p)

 Inputs:
 J1   tapered fourier transform in condition 1
 J2   tapered fourier transform in condition 2
 p      p value for test (default: 0.05)
 plt    'y' for plot and 'n' for no plot
 f      frequencies (useful for plotting)


 Dimensions: J1: frequencies x number of samples in condition 1
             J2: frequencies x number of samples in condition 2
              number of samples = number of trials x number of tapers
 Outputs:
 dz    test statistic (will be distributed as N(0,1) under H0
 vdz   Arvesen estimate of the variance of dz
 Adz   1/0 for accept/reject null hypothesis of equal population
       coherences based dz ~ N(0,1)
 
 
 Note: all outputs are functions of frequency

 References: Arvesen, Jackkknifing U-statistics, Annals of Mathematical
 Statisitics, vol 40, no. 6, pg 2076-2100 (1969)

0001 function [dz,vdz,Adz]=two_group_test_spectrum(J1,J2,p,plt,f)
0002 % function [dz,vdz,Adz]=two_group_test_spectrum(J1,J2,p,plt,f)
0003 % Test the null hypothesis (H0) that data sets J1, J2 in
0004 % two conditions c1,c2 have equal population spectrum
0005 %
0006 % Usage:
0007 % [dz,vdz,Adz]=two_sample_test_spectrum(J1,J2,p)
0008 %
0009 % Inputs:
0010 % J1   tapered fourier transform in condition 1
0011 % J2   tapered fourier transform in condition 2
0012 % p      p value for test (default: 0.05)
0013 % plt    'y' for plot and 'n' for no plot
0014 % f      frequencies (useful for plotting)
0015 %
0016 %
0017 % Dimensions: J1: frequencies x number of samples in condition 1
0018 %             J2: frequencies x number of samples in condition 2
0019 %              number of samples = number of trials x number of tapers
0020 % Outputs:
0021 % dz    test statistic (will be distributed as N(0,1) under H0
0022 % vdz   Arvesen estimate of the variance of dz
0023 % Adz   1/0 for accept/reject null hypothesis of equal population
0024 %       coherences based dz ~ N(0,1)
0025 %
0026 %
0027 % Note: all outputs are functions of frequency
0028 %
0029 % References: Arvesen, Jackkknifing U-statistics, Annals of Mathematical
0030 % Statisitics, vol 40, no. 6, pg 2076-2100 (1969)
0031 
0032 if nargin < 2; error('Need four sets of Fourier transforms'); end;
0033 if nargin < 4 || isempty(plt); plt='n'; end;
0034 %
0035 % Test for matching dimensionalities
0036 %
0037 m1=size(J1,2); % number of samples, condition 1
0038 m2=size(J2,2); % number of samples, condition 2
0039 dof1=m1; % degrees of freedom, condition 1
0040 dof2=m2; % degrees of freedom, condition 2
0041 if nargin < 5 || isempty(f); f=size(J1,1); end;
0042 if nargin < 3 || isempty(p); p=0.05; end; % set the default p value
0043 
0044 %
0045 % Compute the individual condition spectra, coherences
0046 %
0047 S1=conj(J1).*J1; % spectrum, condition 1
0048 S2=conj(J2).*J2; % spectrum, condition 2
0049 
0050 Sm1=squeeze(mean(S1,2)); % mean spectrum, condition 1
0051 Sm2=squeeze(mean(S2,2)); % mean spectrum, condition 2
0052 %
0053 % Compute the statistic dz, and the probability of observing the value dz
0054 % given an N(0,1) distribution i.e. under the null hypothesis
0055 %
0056 bias1=psi(dof1)-log(dof1); bias2=psi(dof2)-log(dof2); % bias from Thomson & Chave
0057 var1=psi(1,dof1); var2=psi(1,dof2); % variance from Thomson & Chave
0058 z1=log(Sm1)-bias1; % Bias-corrected Fisher z, condition 1
0059 z2=log(Sm2)-bias2; % Bias-corrected Fisher z, condition 2
0060 dz=(z1-z2)/sqrt(var1+var2); % z statistic
0061 
0062 % Bug fix
0063 % pdz=normpdf(dz,0,1); % probability of observing value dz
0064 pdz = 2*normcdf(-abs(dz),0,1); % probability of observing value dz
0065 
0066 %
0067 % The remaining portion of the program computes Jackknife estimates of the mean (mdz) and variance (vdz) of dz
0068 %
0069 samples1=[1:m1];
0070 samples2=[1:m2];
0071 %
0072 % Leave one out of one sample
0073 %
0074 bias11=psi(dof1-1)-log(dof1-1); var11=psi(1,dof1-1);
0075 for i=1:m1;
0076     ikeep=setdiff(samples1,i); % all samples except i
0077     Sm1=squeeze(mean(S1(:,ikeep),2)); % 1 drop mean spectrum, data 1, condition 1
0078     z1i(:,i)=log(Sm1)-bias11; % 1 drop, bias-corrected Fisher z, condition 1
0079     dz1i(:,i)=(z1i(:,i)-z2)/sqrt(var11+var2); % 1 drop, z statistic, condition 1
0080     ps1(:,i)=m1*dz-(m1-1)*dz1i(:,i);
0081 end; 
0082 ps1m=mean(ps1,2);
0083 bias21=psi(dof2-1)-log(dof2-1); var21=psi(1,dof2-1);
0084 for j=1:m2;
0085     jkeep=setdiff(samples2,j); % all samples except j
0086     Sm2=squeeze(mean(S2(:,jkeep),2)); % 1 drop mean spectrum, data 2, condition 2
0087     z2j(:,j)=log(Sm2)-bias21; % 1 drop, bias-corrected Fisher z, condition 2
0088     dz2j(:,j)=(z1-z2j(:,j))/sqrt(var1+var21); % 1 drop, z statistic, condition 2
0089     ps2(:,j)=m2*dz-(m2-1)*dz2j(:,j);
0090 end;
0091 %
0092 % Leave one out, both samples
0093 % and pseudo values
0094 % for i=1:m1;
0095 %     for j=1:m2;
0096 %         dzij(:,i,j)=(z1i(:,i)-z2j(:,j))/sqrt(var11+var21);
0097 %         dzpseudoval(:,i,j)=m1*m2*dz-(m1-1)*m2*dz1i(:,i)-m1*(m2-1)*dz2j(:,j)+(m1-1)*(m2-1)*dzij(:,i,j);
0098 %     end;
0099 % end;
0100 %
0101 % Jackknife mean and variance
0102 %
0103 % dzah=sum(sum(dzpseudoval,3),2)/(m1*m2);
0104 ps2m=mean(ps2,2);
0105 % dzar=(sum(ps1,2)+sum(ps2,2))/(m1+m2);
0106 vdz=sum((ps1-ps1m(:,ones(1,m1))).*(ps1-ps1m(:,ones(1,m1))),2)/(m1*(m1-1))+sum((ps2-ps2m(:,ones(1,m2))).*(ps2-ps2m(:,ones(1,m2))),2)/(m2*(m2-1));
0107 % vdzah=sum(sum((dzpseudoval-dzah(:,ones(1,m1),ones(1,m2))).*(dzpseudoval-dzah(:,ones(1,m1),ones(1,m2))),3),2)/(m1*m2);
0108 %
0109 % Test whether H0 is accepted at the specified p value
0110 %
0111 Adz=zeros(size(dz));
0112 x=norminv([p/2 1-p/2],0,1);
0113 indx=find(dz>=x(1) & dz<=x(2)); 
0114 Adz(indx)=1;
0115 
0116 % Adzar=zeros(size(dzar));
0117 % indx=find(dzar>=x(1) & dzar<=x(2));
0118 % Adzar(indx)=1;
0119 %
0120 % Adzah=zeros(size(dzah));
0121 % indx=find(dzah>=x(1) & dzah<=x(2));
0122 % Adzah(indx)=1;
0123 if strcmp(plt,'y');
0124     if isempty(f) || nargin < 5;
0125         f=linspace(0,1,length(dz));
0126     end;
0127     %
0128     % Compute the coherences
0129     %
0130     S1=squeeze(mean(conj(J1).*J1,2));
0131     S2=squeeze(mean(conj(J2).*J2,2));
0132     %
0133     % Plot the coherence
0134     %
0135     subplot(311); 
0136     plot(f,S1,f,S2); legend('Data 1','Data 2');
0137     set(gca,'FontName','Times New Roman','Fontsize', 16);
0138     ylabel('Spectra');
0139     title('Two group test for spectrum');
0140     subplot(312);
0141     plot(f,dz);
0142     set(gca,'FontName','Times New Roman','Fontsize', 16);
0143     ylabel('Test statistic');
0144     conf=norminv(1-p/2,0,1);
0145     line(get(gca,'xlim'),[conf conf]);
0146     line(get(gca,'xlim'),[-conf -conf]);
0147     subplot(313);
0148     plot(f,vdz);
0149     set(gca,'FontName','Times New Roman','Fontsize', 16);
0150     xlabel('frequency'); ylabel('Jackknifed variance');
0151 end;