FitDatabase/fdb_predictors.m at master · clemenskreutz/FitDatabase · GitHub

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
% fdb = fdb_predictors(fdb)
%
% This function generates a predictor matrix nID x nConfigFields

function fdb = fdb_predictors(fdb)

pred = struct;
pred.ynames = fdb.ID;
pred.model_names = fdb_replace_names(fdb.name)';

try

    for i=1:fdb.info.N
        iy = 0;

        iy = iy+1;
        pred.X{i,iy} = pred.model_names{i};
        if i==1
            pred.xnames{iy} = 'name';
        end

        iy = iy+1;
        pred.X{i,iy} = fdb.checksum.fkt{i};
        if i==1
            pred.xnames{iy} = 'fkt';
        end

        iy = iy+1;
        pred.X{i,iy} = fdb.checksum.data{i};
        if i==1
            pred.xnames{iy} = 'data';
        end

        for j=1:length(fdb.info.fields.config)
            iy = iy+1;
            field = fdb.info.fields.config{j};
            val = fdb.checksum.config{i}.(field);
            pred.X{i,iy} = val2pred(val);
            if i==1
                pred.xnames{iy} = ['config_',field];
            end
        end

        for j=1:length(fdb.info.fields.optim)
            iy = iy+1;
            field = fdb.info.fields.optim{j};
            val = fdb.checksum.optim{i}.(field);
            pred.X{i,iy} = val2pred(val);
            if i==1
                pred.xnames{iy} = ['optim_',field];
            end
        end

        for j=1:length(fdb.info.fields.para)
            iy = iy+1;
            field = fdb.info.fields.para{j};
            val = fdb.checksum.para{i}.(field);

            pred.X{i,iy} = val2pred(val);
            if i==1
                pred.xnames{iy} = ['para_',field];
            end
        end
    end

catch ERR
    [i,iy]
    rethrow(ERR)
end

anz = NaN(1,size(pred.X,2));
for i=1:size(pred.X,2)
    anz(i) = length(unique(pred.X(:,i)));
end

pred.X = pred.X(:,anz>1);
pred.xnames = pred.xnames(anz>1);


fdb.predictors = pred;
fdb = fdb_levels(fdb);


function fdb = fdb_levels(fdb)
X = fdb.predictors.X;
xnames = fdb.predictors.xnames;
model_names = fdb.predictors.model_names;
model_lev = levels(model_names);

fdb.predictors.all = struct;
fdb.predictors.all.xlevels = cell(size(xnames));
fdb.predictors.all.replicates = cell(size(xnames));
fdb.predictors.all.default = struct;

for i=1:size(X,2)
    [l,anz] = levels(X(:,i));
    [~,rf] = sort(-anz);
    l = l(rf);
    anz = anz(rf);

    fdb.predictors.all.xlevels{i} = l;
    fdb.predictors.all.replicates{i} = anz;
    fdb.predictors.all.default.(xnames{i}) = l{1};

    for n=1:length(model_lev)
        [~,ind] = intersect(model_names,model_lev{n});

        [l,anz] = levels(X(ind,i));
        [~,rf] = sort(-anz);
        l = l(rf);
        anz = anz(rf);

        fdb.predictors.(model_lev{n}).xlevels{i} = l;
        fdb.predictors.(model_lev{n}).replicates{i} = anz;
        fdb.predictors.(model_lev{n}).default.(xnames{i}) = l{1};
    end
end

fdb.info.predictor_status = 1; % predictors now up-to-date


function predval = val2pred(val)
if isnan(val)
    predval = 'NaN';
elseif isempty(val)
    predval = 'NA';
elseif isnumeric(val)
    predval = sprintf('%d',val);
elseif islogical(val)
    predval = sprintf('%i',val);
elseif ischar(val)
    predval = val;
else
    predval = char(string(val));
end