Homec4science
Diffusion Research Scripts (Thomas Bolton) (master) Browse

Scripts_Sharing/ISFC/Utils/FindExcursions.m
bdfe7aadccccmaster

Last Change
Raw File

FindExcursions.m
View Options

%% This function computes a distribution of values for the considered
% null data (for each connection pair), and determines (in
% two-tailed fashion) the thresholds past which an excursion of
% connectivity in the data is significant. Then, those moments are returned
%
% Inputs:
%
% - Null and Data are the null and the actual data (cell arrays n_subject
% with size n_pairs x n_TP)
% - percentile (in percent) is the level of significance (typically 2.5)
function [Out] = FindExcursions(Null,Data,percentile,Method_type,PureRS)
if(size(Data,1) ~= size(Null,1))
errordlg('The null data and the actual data do not have the same dimensions !!');
end
% Number of region pairs
n_pairs = size(Data{1},1);
n_subjects = length(Data);
switch Method_type
% In this case, we want to use the null from the same subject as
% the actual data to threshold
case 'Subject-Wise'
for s = 1:n_subjects
% Determines the percentiles along the time dimension, so PCT is a
% n_pair x 2 vector
PCT{s} = prctile(Null{s},[100-percentile, percentile],2);
end
% In this case, we want to marginalize the distributions across
% subjects
case 'Subject-Marginalized'
% Aggregates the data of all subjects
Null_marginalized = [];
for s = 1:n_subjects
Null_marginalized = [Null_marginalized,Null{s}];
end
disp(['Number of datapoints: ',num2str(size(Null_marginalized,2))]);
% Does the percentile computations
for s = 1:n_subjects
PCT{s} = prctile(Null_marginalized,[100-percentile, percentile],2);
end
case 'Subject-Marginalized-Added-RS'
% Aggregates the data of all subjects
Null_marginalized = [];
for s = 1:n_subjects
Null_marginalized = [Null_marginalized,Null{s}];
end
n_RS_subjects = length(PureRS);
for s = 1:n_RS_subjects
Null_marginalized = [Null_marginalized,PureRS{s}];
end
disp(['Number of datapoints: ',num2str(size(Null_marginalized,2))]);
% Does the percentile computations
for s = 1:n_subjects
PCT{s} = prctile(Null_marginalized,[100-percentile, percentile],2);
end
case 'Gaussian'
Mu = nan(n_pairs,1);
Sigma = nan(n_pairs,1);
for i = 1:n_pairs
disp(['Pair ',num2str(i),'...']);
% Samples the current null time course of interest
tmp_TC = Null(i,:);
% Computes the distribution of values
[Counts,Xaxis] = hist(tmp_TC,-1:0.1:1);
% Performs the fitting (Gaussian)
tmp_fit = fit(Xaxis',Counts','gauss1');
% Mean and variance
Mu(i) = tmp_fit.b1;
Sigma(i) = sqrt(tmp_fit.c1/2);
end
PCT(:,1) = norminv(100-percentile,Mu,Sigma);
PCT(:,2) = norminv(percentile,Mu,Sigma);
end
Out = zeros(n_pairs,size(Data{s},2),n_subjects);
for s = 1:n_subjects
for i = 1:n_pairs
Out(i,Data{s}(i,:)>PCT{s}(i,2),s) = 1;
Out(i,Data{s}(i,:)<PCT{s}(i,1),s) = -1;
end
end
end
c4science · Help