-
Notifications
You must be signed in to change notification settings - Fork 1
/
hgf_fit_data.m
207 lines (178 loc) · 8.21 KB
/
hgf_fit_data.m
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
%importing all the responses per each subgroup
response_user_none = importdata('hgf_input\rules_for_response_none.csv');
response_user_mild = importdata('hgf_input\rules_for_response_mild.csv');
response_user_moderated = importdata('hgf_input\rules_for_response_moderated.csv');
response_user_moderatelysevere= importdata('hgf_input\rules_for_response_moderatelysevere.csv');
%loading our input responses to the hgf
u = load('hgf_input\input_hgf.txt');
%% First step is to obtain the posteriors to use as priors for fitting each individual
est = tapas_fitModel([],...
u,...
'tapas_hgf_categorical_config',...
'tapas_bayes_optimal_categorical_config',...
'tapas_quasinewton_optim_config')
%plotting ideal response to the set shifting task
tapas_hgf_categorical_plotTraj(est)
%plotting the correlation plot between the parameters
tapas_fit_plotCorr(est)
%% part 2 - fitting subgroups
%use as priors the parameters obtained in step1 that have been stored in
%the file 'tapas_hgf_categorical_config_game'
%% A) Fit No depression with general priors
for i = 1:size(response_user_none,2)
est_none(i) = tapas_fitModel(u,...
response_user_none(2:end,i),...
'tapas_hgf_categorical_config_game',...
'tapas_bayes_optimal_categorical_config',...
'tapas_quasinewton_optim_config')
end
%% B) Plot all participants - no depression subgroup
for i = 1:39
tapas_hgf_categorical_plotTraj(est_none(i))
end
%% C) Calculate and plot the bayesian average - no depression subgroup
bpa_none = tapas_bayesian_parameter_average(est_none(1),est_none(2),est_none(3),est_none(4),est_none(5),est_none(6),est_none(7),est_none(8),est_none(9),est_none(10),est_none(11),est_none(12),est_none(13),est_none(14),est_none(15),est_none(16),est_none(17),est_none(18),est_none(19),est_none(20),est_none(21),est_none(22),est_none(23), est_none(24),est_none(25), est_none(26),est_none(27),est_none(28),est_none(29),est_none(30),est_none(31),est_none(32),est_none(35),est_none(36),est_none(37),est_none(38),est_none(39));
tapas_hgf_categorical_plotTraj(bpa_none)
tapas_fit_plotCorr(bpa_none)
%% D) Extract parameters - no depression subgroup
k_none = zeros(39,1);
for i = 1:39
k_none(i)=est_none(i).p_prc.ka;
end
writematrix(k_none,'kappa_none.csv')
om_none = zeros(39,1);
for i = 1:39
om_none(i)=est_none(i).p_prc.om;
end
writematrix(om_none,'om_none.csv')
th_none = zeros(39,1);
for i = 1:39
th_none(i)=est_none(i).p_prc.th;
end
writematrix(th_none,'none.csv')
%% A) Fitting Mild Depression with general prior
for i = 1:size(response_user_mild,2)
est_mild_g(i) = tapas_fitModel(u,...
response_user_mild(2:end,i),...
'tapas_hgf_categorical_config_game',...
'tapas_bayes_optimal_categorical_config',...
'tapas_quasinewton_optim_config')
end
%% %% B) Plot all participants - mild depression subgroup
for i = 1:size(response_user_mild,2)
tapas_hgf_categorical_plotTraj(est_mild_g(i))
end
%% C) calculate and plot ayesian Parameter Averaging - Mild depression subgroup
bpa_mild_g = tapas_bayesian_parameter_average(est_mild_g(1),est_mild_g(2),est_mild_g(3),est_mild_g(4),est_mild_g(5),est_mild_g(6),est_mild_g(7),est_mild_g(8),est_mild_g(10),est_mild_g(11),est_mild_g(12),est_mild_g(13),est_mild_g(14),est_mild_g(15),est_mild_g(16),est_mild_g(17),est_mild_g(18),est_mild_g(19),est_mild_g(20),est_mild_g(21),est_mild_g(22),est_mild_g(23),est_mild_g(24),est_mild_g(25),est_mild_g(26),est_mild_g(27),est_mild_g(28),est_mild_g(29),est_mild_g(30));
tapas_hgf_categorical_plotTraj(bpa_mild_g)
tapas_fit_plotCorr(bpa_mild_g)
%% D) Extract parameters per person - Mild depression subgroup
k_mild = zeros(30,1);
for i = 1:30
k_mild(i)=est_mild_g(i).p_prc.ka;
end
writematrix(k_mild,'k_mild.csv')
om_mild = zeros(30,1);
for i = 1:30
om_mild(i)=est_mild_g(i).p_prc.om;
end
writematrix(om_mild,'om_mild.csv')
th_mild = zeros(30,1);
for i = 1:30
th_mild(i)=est_mild_g(i).p_prc.th;
end
writematrix(th_mild,'th_mild.csv')
%% A) Fitting Moderated depression participants with general prior
for i = 1:size(response_user_moderated,2)
est_moderated_g(i) = tapas_fitModel(u,...
response_user_moderated(2:end,i),...
'tapas_hgf_categorical_config_game',...
'tapas_bayes_optimal_categorical_config',...
'tapas_quasinewton_optim_config')
end
%% B) Plot all participants - Moderated depression subgroup
for i = 1:10
tapas_hgf_categorical_plotTraj(est_moderated_g(i))
tapas_fit_plotCorr(est_moderated_g(i))
end
%% C) Calculate and plot Bayesian Parameter Averaging - Moderated depression subgroup
bpa_moderateds_g = tapas_bayesian_parameter_average(est_moderated_g(1),est_moderated_g(2),est_moderated_g(3),est_moderated_g(4),est_moderated_g(5),est_moderated_g(6),est_moderated_g(7),est_moderated_g(8),est_moderated_g(10));
tapas_hgf_categorical_plotTraj(bpa_moderateds_g)
tapas_fit_plotCorr(bpa_moderateds_g)
%% D) Extract parameters per person - Moderated depression subgroup
k_moderateds = zeros(10,1);
for i = 1:10
k_moderateds(i)=est_moderated_g(i).p_prc.ka
end
writematrix(k_moderateds,'kappa_modesev.csv')
om_moderateds = zeros(10,1);
for i = 1:10
om_moderateds(i)=est_moderated_g(i).p_prc.om
end
writematrix(om_moderateds,'omega_modesev.csv')
th_moderateds = zeros(10,1);
for i = 1:10
th_moderateds(i)=est_moderated_g(i).p_prc.th
end
writematrix(th_moderateds,'theta_modesev.csv')
%% A) Fit Moderated/Severe depression subgroup with general priors
for i = 1:size(response_user_moderatelysevere,2)
est_moderatelysevere_g(i) = tapas_fitModel(u,...
response_user_moderatelysevere(2:end,i),...
'tapas_hgf_categorical_config_game',...
'tapas_bayes_optimal_categorical_config',...
'tapas_quasinewton_optim_config')
end
%% B) Plot all participants - Moderated/Severe depression subgroup
for i = 1:3
tapas_hgf_categorical_plotTraj(est_moderatelysevere_g(i))
tapas_fit_plotCorr(est_moderatelysevere_g(i))
end
%% C) Bayesian Parameter Averaging - Moderated/Severe depression subgroup
%Average for only three cases across wide PHQ9 scores makes little sense
bpa_moderatelysevere = tapas_bayesian_parameter_average(est_moderatelysevere_g(1), est_moderatelysevere_g(2),est_moderatelysevere_g(3));
tapas_hgf_categorical_plotTraj(bpa_moderatelysevere)
tapas_fit_plotCorr(bpa_moderatelysevere)
%% D) Extract parameters per person - Moderated/Severe
om_moderatelysevere = zeros(3,1);
om_moderatelysevere(1)=est_moderatelysevere_g(1).p_prc.om
om_moderatelysevere(2)=est_moderatelysevere_g(2).p_prc.om
om_moderatelysevere(3)=est_moderatelysevere_g(3).p_prc.om
writematrix(om_moderatelysevere,'omega_moderatelysevere.csv')
th_moderatelysevere = zeros(3,1);
th_moderatelysevere(1)=est_moderatelysevere_g(1).p_prc.th
th_moderatelysevere(2)=est_moderatelysevere_g(2).p_prc.th
th_moderatelysevere(3)=est_moderatelysevere_g(3).p_prc.th
writematrix(th_moderatelysevere,'theta_moderatelysevere.csv')
%%
%Box plots for the bayes_optimal_categorical
%read csv containg all omegas por subgroup
boxnone = readmatrix('mino');
boxmild = readmatrix('modo');
boxmods = readmatrix('mso'); %contains moderaed, moderately severe and severe
g1 = repmat({'None'},size(boxnone,1),1);
g2 = repmat({'Mild'},(size(boxmild,1)),1);
g3 = repmat({'Moderated to severe'},(size(boxmods,1)),1);
g = [g1; g2; g3];
%Boxplot correct
x_correct = [boxnone;boxmild;boxmods];
boxplot(x_correct,g)
title('ω Values')
ylabel('tonic volatility ω')
xlabel('Level of depression')
%%
%Box plots for the softmax
%read csv containg all omegas por subgroup
boxnone2 = readmatrix('mino2');
boxmild2 = readmatrix('modo2');
boxmods2 = readmatrix('mso2'); %contains moderaed, moderately severe and severe
g1 = repmat({'None'},size(boxnone2,1),1);
g2 = repmat({'Mild'},(size(boxmild2,1)),1);
g3 = repmat({'Moderated to severe'},(size(boxmods2,1)),1);
g = [g1; g2; g3];
%Boxplot correct
x_correct = [boxnone2;boxmild2;boxmods2];
boxplot(x_correct,g)
title('ω Values')
ylabel('tonic volatility ω')
xlabel('Level of depression')