add every visit no jump results

src/main/java/core/algo/EpisodicLearning.java

@@ -5,7 +5,7 @@ import core.Environment;
 import core.LearningConfig;
 import core.StepResult;
 import core.listener.LearningListener;
-import core.policy.EpsilonGreedyPolicy;
+import example.DinoSampling;
 import lombok.Getter;
 import lombok.Setter;
 
@@ -104,7 +104,7 @@ public abstract class EpisodicLearning<A extends Enum> extends Learning<A> imple
         timestamp++;
         timestampCurrentEpisode++;
         // TODO: more sophisticated way to check convergence
-        if(timestampCurrentEpisode > 300000){
+        if(timestampCurrentEpisode > 50000) {
             converged = true;
             // t
             File file = new File(DinoSampling.FILE_NAME);

src/main/java/evironment/jumpingDino/DinoWorld.java

@@ -44,7 +44,7 @@ public class DinoWorld implements Environment<DinoAction>, Visualizable {
     @Override
     public StepResultEnvironment step(DinoAction action) {
         boolean done = false;
-        int reward = 0;
+        int reward = 1;
 
         if(action == DinoAction.JUMP){
             dino.jump();
@@ -68,7 +68,7 @@ public class DinoWorld implements Environment<DinoAction>, Visualizable {
             spawnNewObstacle();
         }
         if(ranIntoObstacle()) {
-            reward = -1;
+            reward = 0;
             done = true;
         }
 

src/main/java/example/DinoSampling.java

@@ -3,10 +3,8 @@ package example;
 import core.RNG;
 import core.algo.Method;
 import core.controller.RLController;
-import core.controller.RLControllerGUI;
 import evironment.jumpingDino.DinoAction;
 import evironment.jumpingDino.DinoWorld;
-import evironment.jumpingDino.DinoWorldAdvanced;
 
 import java.io.File;
 import java.io.IOException;
@@ -16,7 +14,7 @@ import java.nio.file.StandardOpenOption;
 
 public class DinoSampling {
     public static final float f =0.05f;
-    public static final String FILE_NAME = "converge.txt";
+    public static final String FILE_NAME = "convergence.txt";
     public static void main(String[] args) {
         File file = new File(FILE_NAME);
         try {
@@ -24,24 +22,24 @@ public class DinoSampling {
         } catch (IOException e) {
             e.printStackTrace();
         }
-        for(float f = 0.05f; f <=1.003 ; f+=0.05f) {
+        for(float f = 0.06f; f <= 1.003; f += 0.05f) {
             try {
                 Files.writeString(Path.of(file.getPath()), f + ",", StandardOpenOption.APPEND);
             } catch (IOException e) {
                 e.printStackTrace();
             }
-            for (int i = 1; i <= 100; i++) {
+            for(int i = 1; i <= 10; i++) {
                 System.out.println("seed: " + i * 13);
-                RNG.setSeed(i * 13);
+                RNG.setSeed(i * 17);
 
                 RLController<DinoAction> rl = new RLController<>(
-                        new DinoWorldAdvanced(),
-                        Method.Q_LEARNING_OFF_POLICY_CONTROL,
+                        new DinoWorld(),
+                        Method.MC_CONTROL_EVERY_VISIT,
                         DinoAction.values());
                 rl.setDelay(0);
-                rl.setDiscountFactor(0.99f);
+                rl.setDiscountFactor(1f);
                 rl.setEpsilon(f);
-                rl.setLearningRate(0.9f);
+                rl.setLearningRate(1f);
                 rl.setNrOfEpisodes(400000);
                 rl.start();
             }