ComputeSurrogatePercentiles.m
e5f5005bb2efmaster

ComputeSurrogatePercentiles.m
View Options

			%% This function computes, for a set of innovation time courses, the
			% percentile values at a given alpha-level (i.e. the threshold past which
			% an innovation in real data is deemed significant)
			%
			% Inputs:
			% - I (n_tp x n_ret_vox) is the matrix of surrogate data innovation signals
			% - alpha is the level (in percentage) at which we want to threshold; it
			% should be a 2-element vector (e.g. [1 99] for the 1st and 99th
			% percentiles)
			% - fid points towards the file where we want to write about the performed
			% steps
			%
			% Outputs:
			% - PC is a (2 x n_ret_vox) matrix with PC(1,:) the lower percentile and
			% PC(2,:) the upper percentile values
			function [PC] = ComputeSurrogatePercentiles(I,param,fid)

			% Computes percentiles for each time course (each column); PC is a 2 x
			% n_ret_vox matrix
			PC = prctile(I,param.alpha);

			% Writes in log-file
			WriteInformation(fid,['Computed voxel-wise percentiles: average values are ',...
			num2str(mean(PC(1,:))),' for bottom, and ',num2str(mean(PC(2,:))),...
			' for top percentiles...']);
			end