@@ -32,6 +32,36 @@ void Agent::initQ() } } +void Agent::changeAlpha(int i) +{ + setAlpha(i/10); +} + +void Agent::changeGamma(int i) +{ + setGamma(i/10); +} + +void Agent::changeEps(int i) +{ + switch(i){ + case 0: + setEps(0); + case 1: + setEps(0.0001); + case 2: + setEps(0.001); + case 3: + setEps(0.01); + case 4: + setEps(0.1); + case 5: + setEps(1.0); + default: + setEps(0); + } +} + Action Agent::selectAction(Status s) { Reward max = -100000000; @@ -47,60 +77,60 @@ Action Agent::selectAction(Status s) } p = m_Q.find(MyQ(s,Action(0,-1))); if(p != m_Q.end()){ - if (p.value() > max) {max = p.value(); a = Action(0,-1);} + if (p.value() > max) {max = p.value(); a.first = 0; a.second = -1;} } p = m_Q.find(MyQ(s,Action(0,1))); if(p != m_Q.end()){ - if (p.value() > max) {max = p.value(); a = Action(0,1);} + if (p.value() > max) {max = p.value(); a.first = 0; a.second = 1;} } p = m_Q.find(MyQ(s,Action(1,1))); if(p != m_Q.end()){ - if (p.value() > max) {max = p.value(); a = Action(1,1);} + if (p.value() > max) {max = p.value(); a.first = 1; a.second = 1;} } p = m_Q.find(MyQ(s,Action(-1,1))); if(p != m_Q.end()){ - if (p.value() > max) {max = p.value(); a = Action(-1,1);} + if (p.value() > max) {max = p.value(); a.first = -1; a.second = 1;} } p = m_Q.find(MyQ(s,Action(1,-1))); if(p != m_Q.end()){ - if (p.value() > max) {max = p.value(); a = Action(1,-1);} + if (p.value() > max) {max = p.value(); a.first = 1; a.second = -1;} } p = m_Q.find(MyQ(s,Action(-1,-1))); if(p != m_Q.end()){ - if (p.value() > max) {max = p.value(); a = Action(-1,-1);} - } - p = m_Q.find(MyQ(s,Action(2,0))); - if(p != m_Q.end()){ - if (p.value() > max) {max = p.value(); a = Action(2,0);} - } - p = m_Q.find(MyQ(s,Action(-2,0))); - if(p != m_Q.end()){ - if (p.value() > max) {max = p.value(); a = Action(-2,0);} - } - p = m_Q.find(MyQ(s,Action(0,-2))); - if(p != m_Q.end()){ - if (p.value() > max) {max = p.value(); a = Action(0,-2);} - } - p = m_Q.find(MyQ(s,Action(0,2))); - if(p != m_Q.end()){ - if (p.value() > max) {max = p.value(); a = Action(0,2);} - } - p = m_Q.find(MyQ(s,Action(2,2))); - if(p != m_Q.end()){ - if (p.value() > max) {max = p.value(); a = Action(2,2);} - } - p = m_Q.find(MyQ(s,Action(-2,2))); - if(p != m_Q.end()){ - if (p.value() > max) {max = p.value(); a = Action(-2,2);} - } - p = m_Q.find(MyQ(s,Action(2,-2))); - if(p != m_Q.end()){ - if (p.value() > max) {max = p.value(); a = Action(2,-2);} - } - p = m_Q.find(MyQ(s,Action(-2,-2))); - if(p != m_Q.end()){ - if (p.value() > max) {max = p.value(); a = Action(-2,-2);} - } + if (p.value() > max) {max = p.value(); a.first = -1; a.second = -1;} + } +// p = m_Q.find(MyQ(s,Action(2,0))); +// if(p != m_Q.end()){ +// if (p.value() > max) {max = p.value(); a = Action(2,0);} +// } +// p = m_Q.find(MyQ(s,Action(-2,0))); +// if(p != m_Q.end()){ +// if (p.value() > max) {max = p.value(); a = Action(-2,0);} +// } +// p = m_Q.find(MyQ(s,Action(0,-2))); +// if(p != m_Q.end()){ +// if (p.value() > max) {max = p.value(); a = Action(0,-2);} +// } +// p = m_Q.find(MyQ(s,Action(0,2))); +// if(p != m_Q.end()){ +// if (p.value() > max) {max = p.value(); a = Action(0,2);} +// } +// p = m_Q.find(MyQ(s,Action(2,2))); +// if(p != m_Q.end()){ +// if (p.value() > max) {max = p.value(); a = Action(2,2);} +// } +// p = m_Q.find(MyQ(s,Action(-2,2))); +// if(p != m_Q.end()){ +// if (p.value() > max) {max = p.value(); a = Action(-2,2);} +// } +// p = m_Q.find(MyQ(s,Action(2,-2))); +// if(p != m_Q.end()){ +// if (p.value() > max) {max = p.value(); a = Action(2,-2);} +// } +// p = m_Q.find(MyQ(s,Action(-2,-2))); +// if(p != m_Q.end()){ +// if (p.value() > max) {max = p.value(); a = Action(-2,-2);} +// } return a; } reinforce/agent.cpp @@ -9,6 +9,9 @@ class Agent : public QObject Q_OBJECT public: Agent(Environment *env, double alpha, double gamma, double eps); + double alpha() { return m_alpha; } + double gamma() { return m_gamma; } + double epsilon() { return m_eps; } void setAlpha(double alpha){ m_alpha = alpha; } void setGamma(double gamma){ m_gamma = gamma; } void setEps(double eps){ m_eps = eps; } @@ -16,6 +19,10 @@ public: void initQ(); bool isFin(Status st); Action selectAction(Status s); +public slots: + void changeAlpha(int i); + void changeGamma(int i); + void changeEps(int i); private: Environment *m_env; double m_alpha; reinforce/agent.h @@ -14,6 +14,7 @@ public: Status getStatus() { return m_cur_status; } signals: void statusChanged(Status s); + private: Status m_cur_status; }; reinforce/environment.h @@ -2,6 +2,8 @@ #include "ui_mainwindow.h" #include "environment.h" +#include <QMessageBox> + MainWindow::MainWindow(QWidget *parent) : QMainWindow(parent), ui(new Ui::MainWindow) { @@ -9,8 +11,17 @@ MainWindow::MainWindow(QWidget *parent) env = new Environment(); agent = new Agent(env, 0.5, 0.5, 0.1); // double alpha, double gamma, double eps + ui->alphaCombo->setCurrentIndex(5); + ui->gammaCombo->setCurrentIndex(5); + ui->epsilonCombo->setCurrentIndex(4); connect(env, SIGNAL(statusChanged(Status)), this, SLOT(recvStatusChanged(Status))); + connect(ui->alphaCombo, SIGNAL(currentIndexChanged(int)), + agent, SLOT(changeAlpha(int))); + connect(ui->gammaCombo, SIGNAL(currentIndexChanged(int)), + agent, SLOT(changeGamma(int))); + connect(ui->epsilonCombo, SIGNAL(currentIndexChanged(int)), + agent, SLOT(changeEps(int))); } MainWindow::~MainWindow() @@ -26,12 +37,16 @@ void MainWindow::recvStatusChanged(Status s) void MainWindow::on_initButton_clicked() { + if(Status(10,10) == Status(10,10)){ + QMessageBox::information(NULL, "", "true"); + } + } void MainWindow::on_startButton_clicked() { agent->initQ(); - for(int i=0;i<100;i++){ + for(int i=0;i<3;i++){ agent->doEpisode(); } } reinforce/mainwindow.cpp @@ -227,6 +227,11 @@ <string>0.1</string> </property> </item> + <item> + <property name="text"> + <string>1.0</string> + </property> + </item> </widget> </item> </layout> reinforce/mainwindow.ui 84bd05d69c64f92cd7a2f5c44fc8d02ae3ceca0a NOSE Takafumi ahya365@gmail.com http://github.com/plus7/plus7pub/commit/93fdbc5fb07b79ee5672f81cadb79e80e59b3532 93fdbc5fb07b79ee5672f81cadb79e80e59b3532 2009-04-29T20:43:24-07:00 2009-04-29T20:43:24-07:00 nf 54c91070b1582bc26d2d7af9ba0954ccc264d048 NOSE Takafumi ahya365@gmail.com