ORIGINAL_MAIN

[duriseul]

clear; clc; rng(44501)

NumModel = 2;

load('../Mydata_Orderblk.mat') numsub = size(Mydata_Orderblk,1); f = fieldnames(Mydata_Orderblk{1,1}); maxB = 20; ctr = 1;

v = 6.25; parameters = randn(1,2); prior_RL = struct('mean',zeros(2,1),'variance',v); % note dimension of 'mean'

fname_RL_8S_v2 = 'lap_RL_8S_v2.mat'; fname_RL_4S_v2 = 'lap_RL_4S_v2.mat';

fcbm_maps = {'lap_RL_8S_v2.mat', 'lap_RL_4S_v2.mat'}; models = {@model_RL_8states_v2, @model_RL_4states_v2};

MF = zeros(maxB, NumModel); MFall = zeros(maxB, numsub, NumModel); NLL = zeros(maxB, NumModel, numsub); Theta = zeros(maxB, NumModel, numsub, 2);

for tr=1: maxB ctr = 1; Data= []; for j=1:numsub subj = Mydata_Orderblk{j,tr}; if isempty(subj) else Data{ctr, 1}=subj; ctr = ctr+1; end end

CBM8S = cbm_lap(Data, @model_RL_8states_v2, prior_RL, fname_RL_8S_v2);
CBM4S = cbm_lap(Data, @model_RL_4states_v2, prior_RL, fname_RL_4S_v2);

fname_hbi_blk = ['hbi_RLv2_8-4S_blk_',num2str(tr),'.mat'];

cbm_hbi(Data,models,fcbm_maps,fname_hbi_blk);
load(fname_hbi_blk);

MF(tr, :)=cbm.output.model_frequency;

% get indices of subjects for which there was data for this block
subidx = ~cellfun(@isempty, Mydata_Orderblk(:,tr));

MFall(tr, subidx, :)=cbm.output.responsibility;

NLL(tr,1,subidx) = -CBM8S.math.loglik;
NLL(tr,2,subidx) = -CBM4S.math.loglik;

Theta(tr,1,subidx,:) = [CBM8S.math.theta{:}]';
Theta(tr,2,subidx,:) = [CBM4S.math.theta{:}]';

end save('Mf.mat', 'MF','MFall') save('NLL.mat','NLL') save('Theta.mat','Theta')

delete hbi lap