Implementation for Neural Net menu.

* Save best neural net.
* Load best neural net (e.g. from some earlier execution).
* Override all neural nets using the current best neural net.

src/main/java/thommynator/App.java

@@ -1,21 +1,32 @@
 package thommynator;
 
+import lombok.extern.slf4j.Slf4j;
 import thommynator.game.GameCanvas;
 import thommynator.game.Population;
 import thommynator.game.Racetrack;
 
 import javax.swing.*;
 import java.awt.*;
+import java.awt.event.ActionEvent;
+import java.awt.event.ActionListener;
 import java.util.concurrent.atomic.AtomicInteger;
 
-public class App extends JFrame {
+import static thommynator.game.MenuContent.NEURALNET_LOAD;
+import static thommynator.game.MenuContent.NEURALNET_OVERRIDE_BEST;
+import static thommynator.game.MenuContent.NEURALNET_SAVE;
+import static thommynator.game.MenuContent.RACE_START;
+import static thommynator.game.MenuContent.RACE_STOP;
+
+@Slf4j
+public class App extends JFrame implements ActionListener {
 
     public static final int MAP_WIDTH = 800;
     public static final int MAP_HEIGHT = 600;
+    public static final AtomicInteger CAR_ID_FACTORY = new AtomicInteger(0);
     public static final Racetrack RACETRACK = new Racetrack();
+    public static final Population POPULATION = new Population(200);
     public static final int INITIAL_X = 30;
     public static final int INITIAL_Y = 30;
-    public static final AtomicInteger CAR_ID_FACTORY = new AtomicInteger(0);
 
 
     public App() {
@@ -33,20 +44,38 @@ private void initUI() {
 
         // add a menu at the top
         JMenuBar menuBar = new JMenuBar();
+        JMenuItem menuItem = null;
         JMenu raceMenu = new JMenu("Race");
-        raceMenu.add(new JMenuItem("Start"));
-        raceMenu.add(new JMenuItem("Stop"));
+        menuItem = new JMenuItem(RACE_START.getMsg());
+        menuItem.addActionListener(this);
+        raceMenu.add(menuItem);
+
+        menuItem = new JMenuItem(RACE_STOP.getMsg());
+        menuItem.addActionListener(this);
+        raceMenu.add(menuItem);
         menuBar.add(raceMenu);
+
+
         JMenu neuralNetMenu = new JMenu("Neural Net");
-        neuralNetMenu.add(new JMenuItem("Save best..."));
-        neuralNetMenu.add(new JMenuItem("Load best..."));
+        menuItem = new JMenuItem(NEURALNET_SAVE.getMsg());
+        menuItem.addActionListener(this);
+        neuralNetMenu.add(menuItem);
+
+        menuItem = new JMenuItem(NEURALNET_LOAD.getMsg());
+        menuItem.addActionListener(this);
+        neuralNetMenu.add(menuItem);
+
+        menuItem = new JMenuItem(NEURALNET_OVERRIDE_BEST.getMsg());
+        menuItem.addActionListener(this);
+        neuralNetMenu.add(menuItem);
+
         menuBar.add(neuralNetMenu);
         setJMenuBar(menuBar);
 
         // add the content panel
         JPanel contentPanel = new JPanel();
         contentPanel.setLayout(null);
-        contentPanel.add(new GameCanvas(new Population(200)));
+        contentPanel.add(new GameCanvas(POPULATION));
         contentPanel.add(new Racetrack());
 
         setDefaultCloseOperation(JFrame.EXIT_ON_CLOSE);
@@ -58,4 +87,24 @@ private void initUI() {
         // close window
         this.setDefaultCloseOperation(JFrame.EXIT_ON_CLOSE);
     }
+
+    @Override
+    public void actionPerformed(ActionEvent e) {
+
+        String actionCommand = e.getActionCommand();
+        if (actionCommand.equals(NEURALNET_SAVE.getMsg())) {
+            POPULATION.getBestCar().getNeuralNet().save();
+            log.info("Saved best Neural Net.");
+        }
+
+        if (actionCommand.equals(NEURALNET_LOAD.getMsg())) {
+            POPULATION.overrideAllWithJson();
+            log.info("Loaded best Neural Net.");
+        }
+
+        if (actionCommand.equals(NEURALNET_OVERRIDE_BEST.getMsg())) {
+            POPULATION.overrideAllWithBest();
+            log.info("Use the best neural net for all.");
+        }
+    }
 }

src/main/java/thommynator/game/Car.java

@@ -193,4 +193,4 @@ private boolean isInsideCanvas(double x, double y) {
                 && y > buffer
                 && y < (App.MAP_HEIGHT - buffer);
     }
-}
+}

src/main/java/thommynator/game/GameCanvas.java

@@ -81,13 +81,11 @@ public void run() {
 
             long now = System.currentTimeMillis();
             if (now - epochStartTime > MAX_EPOCH_TIME) {
-                population.getBestCar().getNeuralNet().save();
                 population.nextGeneration();
                 epochStartTime = now;
             }
 
             if (!population.isAlive()) {
-                population.getBestCar().getNeuralNet().save();
                 population.nextGeneration();
                 epochStartTime = now;
             }

src/main/java/thommynator/game/MenuContent.java

@@ -0,0 +1,17 @@
+package thommynator.game;
+
+import lombok.AllArgsConstructor;
+import lombok.Getter;
+
+@AllArgsConstructor
+public enum MenuContent {
+
+    RACE_START("Start"),
+    RACE_STOP("Stop"),
+    NEURALNET_SAVE("Save best..."),
+    NEURALNET_LOAD("Load..."),
+    NEURALNET_OVERRIDE_BEST("Override all...");
+
+    @Getter
+    private final String msg;
+}

src/main/java/thommynator/game/Population.java

@@ -18,7 +18,7 @@ public class Population {
     private ArrayList<Car> cars;
 
     public Population(int amountOfCars) {
-        log.debug("Create a new population with {} cars.", amountOfCars);
+        log.debug("Create a new POPULATION with {} cars.", amountOfCars);
         this.amountOfCars = amountOfCars;
 
         cars = new ArrayList<>();
@@ -28,14 +28,14 @@ public Population(int amountOfCars) {
     }
 
     /**
-     * Generate a new population of cars from the current one.
+     * Generate a new POPULATION of cars from the current one.
      * Cars with a high {@link Car#fitness} have a better chance to survive.
      */
     public void nextGeneration() {
         log.debug("Creating new generation...");
         ArrayList<Car> children = new ArrayList<>();
 
-        // make sure, that the best car is for sure in the new population
+        // make sure, that the best car is for sure in the new POPULATION
         Car bestCar = this.getBestCar();
 
         Car child = this.generateChild(bestCar);
@@ -58,7 +58,7 @@ public void nextGeneration() {
     }
 
     /**
-     * Finds the best car of this population and returns it.
+     * Finds the best car of this POPULATION and returns it.
      * The best car is the one with the highest {@link Car#fitness}.
      *
      * @return the {@link Car} with the hightest fitness.
@@ -85,7 +85,7 @@ public Car getBestCar() {
 
     /**
      * Checks if the populations is alive or dead.
-     * A population is dead, if all if its cars are dead. As long as one single car is alive, the whole populations
+     * A POPULATION is dead, if all if its cars are dead. As long as one single car is alive, the whole populations
      * is alive.
      *
      * @return true if alive, otherwise false.
@@ -111,14 +111,14 @@ private Car mutateChild(Car child) {
     }
 
     /**
-     * Calls for all cars in the population the update method {@link Car#updateState()}.
+     * Calls for all cars in the POPULATION the update method {@link Car#updateState()}.
      */
     public void update() {
         cars.parallelStream().forEach(Car::updateState);
     }
 
     /**
-     * Loads the {@link NeuralNet} of the best car and uses it for the whole population.
+     * Loads the {@link NeuralNet} of the best car and uses it for the whole POPULATION.
      * Every {@link Car} will have the same {@link NeuralNet} after this.
      */
     public void overrideAllWithBest() {
@@ -127,19 +127,19 @@ public void overrideAllWithBest() {
     }
 
     /**
-     * Loads a {@link NeuralNet} from a json file and uses it for the whole population.
+     * Loads a {@link NeuralNet} from a json file and uses it for the whole POPULATION.
      * Every {@link Car} will have the same {@link NeuralNet} after this.
      */
     public void overrideAllWithJson() {
         this.overrideAllWithJson("neural-net.json");
     }
 
-    protected void overrideAllWithJson(String fileName) {
+    public void overrideAllWithJson(String fileName) {
         NeuralNet neuralNet = NeuralNet.load(fileName);
         this.overrideAllNeuralNets(neuralNet);
     }
 
-    private void overrideAllNeuralNets(NeuralNet neuralNet) {
+    public void overrideAllNeuralNets(NeuralNet neuralNet) {
         cars.parallelStream().forEach(car -> car.setNeuralNet(new NeuralNet(neuralNet)));
     }