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
|
% data generation
clear all
close all
ddd = datestr(now);
ddd([12 15 18])='-' ;
fnd=[ddd '.txt'];
diary(fnd)
% Metaparams
hyp = 'mar';
ntests = 2; % change it !
nessais = 2; % change it !
m = 1000; % change it !
% INIT
missing = 2;
misv = -9999;
%test = zeros(nessais,6);
test = zeros(nessais,ntests);
taille = zeros(nessais,ntests);
kiki = zeros(nessais,1);
%%%%%%%%%% TESTS %%%%%%%%%%%
for essai=1:nessais
N = ceil(rand*11) + 2
base_proba = .15 + rand/4;
fan_in = ceil(rand*4)
dag=mk_rnd_dag(N,fan_in)
discrete=1:N;
for i=1:N, node_sizes(i) = ceil(rand*4)+1; end
bnet = mk_bnet(dag, node_sizes, discrete);
node_sizes
for i=1:N, bnet.CPD{i} = tabular_CPD(bnet, i); end
if prod([hyp(1:3) == 'mca']+0),
bnet_miss = gener_MCAR_net(bnet, base_proba);
disp(' - MCAR net generated');
elseif prod([hyp(1:3) == 'mar']+0),
bnet_miss = gener_MAR_net(bnet, base_proba);
disp(' - MAR net generated');
end
[data, comp_data, bnet_miss, taux, bnet_orig, notok, dT] = gener_data_from_bnet_miss(bnet_miss, m, base_proba ,1);
base_proba
kiki(essai) = 1-chi2cdf(dT,1);
%test(essai,1)=1-chi2cdf(d,df);
node0=[]; param0=[];
for i=1:ntests
% choose a node
node=ceil(rand*N);
while ~isempty(myintersect(node,node0)) & i<N-2,
node = ceil(rand*N);
end
node0 = myunion(node0,node);
% choose a param
par = parents(bnet_miss.dag, node+N);
sorted=[]; ord=[];
for par1=1:length(par)
ord(par1)=find(par(par1)==bnet_miss.order);
end
[tmp, sorted] = sort(ord);
family = par(sorted);
state = []; stateM = [];
for par1=1:length(family)
state(par1) = ceil(rand*bnet_miss.node_sizes(family(par1)));
if state(par1) == bnet_miss.node_sizes(family(par1)), stateM(par1) = misv; else, stateM(par1) = state(par1); end
end
if prod([hyp(1:3) == 'mca']+0),
family = family-N;
elseif prod([hyp(1:3) == 'mar']+0),
family = family-2*N;
end
family(end+1)=node;
state(end+1) = missing; stateM(end+1)=misv;
CPT = CPT_from_bnet(bnet_miss);
value = CPT{node+N}(subv2ind(bnet_miss.node_sizes(family), state))
% test this param "value"
datamat = bnt_to_mat(data(family,:), misv);
indj=1:m;
for par1 = 1:(length(family)-1)
[tmp, indj] = find(datamat(par1,indj)==stateM(par1));
end
mtest = length(indj);
[tmp, indj] = find(datamat(length(family),indj)==stateM(length(family)));
toto=mtest; nbr_miss=length(indj);
d = ((nbr_miss-value*toto)^2)/(value*toto)...
+ (((toto-nbr_miss)-(1-value)*toto)^2)/((1-value)*toto)
family
%df = prod(bnet_miss.node_sizes(family)-1);
df = 1; % present or missing
%for chiv=1:nchi,
% if d>chi2_table(chi2(chiv),df), test(essai,i,chiv) = 1; end
%end
cumchi=chi2cdf(d,df)
test(essai,i) = 1-cumchi;
taille(essai,i) = toto;
%fprintf('%2.1f%% OF MISSING DATA (%2.1f%%, Khi2 : %2.1f > %2.1f)\n', round(base_proba*10000)/100, round(taux*10000)/100, d, chi2(nchi));
end
clear bnet_miss bnet datamat data node0 family node_sizes CPT indj tmp
end
kiki
taille
test
%eval(['save resgener' ddd 'kiki taille test hyp ntests nessais = 50 m'])
diary off
|
