RSCHMM_ComputeWZ_Optimized.m
c5f5e4464253master

RSCHMM_ComputeWZ_Optimized.m
View Options

			%% This function computes the W and Z parameters present in the likelihood
			%
			% Inputs:
			% - Alpha has size R x 1 and contains the current estimates of intrinsic
			% transition probabilities for the regions
			% - Beta has size R x R, like Gamma, and they contain all the coupling
			% coefficients
			% - X matrices contain the activation states of all but ROI at t and t+1
			% - Y determines whether we have a change in state or not (1 or 0) from t
			% to t+1
			function [W,Z] = RSCHMM_ComputeWZ_Optimized(Alpha,Beta,Gamma,X_causal,X_coact,Y)

			p = cellfun(@(x,y) RSCHMM_Eval_Logistic_Optimized(Alpha,...
			Beta,Gamma,x,y),X_causal,X_coact,'UniformOutput',false);

			Z = cellfun(@(a,b,c,d) Alpha+aBeta+bGamma+(c-d)./(realmin+d.*(1-d)),...
			X_causal,X_coact,Y,p,'UniformOutput',false);

			W = cellfun(@(x) x.*(1-x),p,'UniformOutput',false);
			end