2048qt/CPerceptron.cpp at master · clebail/2048qt · 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
138
139
#include "CPerceptron.h"

CPerceptron::CPerceptron(const QList<int>& nbInputs, int nbNeurone, bool useBiais) {
    QList<CNeurone *> layer;
    int i;
    this->useBiais = useBiais;

    for(i=0;i<nbNeurone;i++) {
        CNeurone *n = new CNeurone(nbInputs.at(i), useBiais);
        layer.append(n);
    }

    layers.append(layer);
}

CPerceptron::~CPerceptron(void) {
    int i;

    for(i=0;i<layers.size();i++) {
        int j;
        QList<CNeurone *> layer = layers.at(i);

        for(j=0;j<layer.size();j++) {
            delete layer.at(j);
        }

        layer.clear();
    }
    layers.clear();
}

void CPerceptron::init(void) {
    int i;

    for(i=0;i<layers.size();i++) {
        int j;

        for(j=0;j<layers.at(i).size();j++) {
            layers.at(i).at(j)->init();
        }
    }
}

QList<double> CPerceptron::eval(const QList<QList<double> >& inputs) {
    QList<double> resultat;
    QList<double> in;
    int i, j;

    for(i=0;i<layers.at(0).size();i++) {
        CNeurone *n = layers.at(0).at(i);
        resultat << n->eval(inputs.at(i));
    }

    for(i=1;i<layers.size();i++) {
        in.clear();
        in << resultat;
        resultat.clear();

        for(j=0;j<layers.at(i).size();j++) {
            CNeurone *n = layers.at(i).at(j);

            resultat << n->eval(in);
        }
    }

    return resultat;
}

void CPerceptron::addLayer(int nbNeurone) {
    int nbInput = layers.last().size();
    int i;
    QList<CNeurone *> layer;

    for(i=0;i<nbNeurone;i++) {
        CNeurone *n = new CNeurone(nbInput, useBiais);
        layer.append(n);
    }

    layers.append(layer);
}

void CPerceptron::from(CPerceptron * p1, CPerceptron * p2) {
    int i;

    for(i=0;i<layers.size();i++) {
        int j;

        for(j=0;j<layers.at(i).size();j++) {
            CNeurone *n = layers.at(i).at(j);
            CNeurone *n1 = p1->layers.at(i).at(j);
            CNeurone *n2 = p2->layers.at(i).at(j);
            int seuil = rand() % n->getNbGene();

            n->from(n1, n2, seuil);
            if(rand() % 100 > 80) {
                n->mute(rand() % n->getNbGene());
            }
        }
    }
}

int CPerceptron::diff(CPerceptron * other) const {
    int i;
    int diff = 0;

    for(i=0;i<layers.size();i++) {
        int j;

        for(j=0;j<layers.at(i).size();j++) {
            CNeurone *n = layers.at(i).at(j);
            CNeurone *no = other->layers.at(i).at(j);
            diff += n->diff(no);
        }
    }

    return diff;
}

QDataStream& operator<<(QDataStream& out, const CPerceptron& perceptron) {
    int i;

    out << (qint32)perceptron.layers.size();
    out << perceptron.useBiais;
    out << (qint32)perceptron.layers.at(0).size();
    for(i=0;i<perceptron.layers.at(0).size();i++) {
        out << perceptron.layers.at(0).at(i);
    }

    for(i=1;i<perceptron.layers.size();i++) {
        int j;

        out << (qint32)perceptron.layers.at(i).size();
        for(j=0;j<perceptron.layers.at(i).size();j++) {
            out << perceptron.layers.at(i).at(j);
        }
    }

    return out;
}