From: charlie <charlie@gerhardus.com>
Date: Sun, 24 Jan 2016 21:12:36 +0000 (+0100)
Subject: cmd options for learner and learned model for minimal input alphabet added as sample.
X-Git-Url: https://git.martlubbers.net/?a=commitdiff_plain;h=0f78791495b0f4f172b02cd14bea0bedcd37f3f2;p=tt2015.git

cmd options for learner and learned model for minimal input alphabet added as sample.
---

diff --git a/a4/tcp/adapter/listener.py b/a4/tcp/adapter/listener.py
index 17d6f21..d64503b 100644
--- a/a4/tcp/adapter/listener.py
+++ b/a4/tcp/adapter/listener.py
@@ -34,7 +34,7 @@ if __name__ == "__main__":
         print 'received: {}'.format(data)
         if data == 'RES':
             print 'resetting the SUT...'
-            sender.sendReset()
+            #sender.sendReset()
             sender = Sender(serverIP="127.0.0.1", networkInterface="lo", isLocal=True, serverPort=serverPort, waitTime=1, isVerbose=0)
             data = ''
             continue
diff --git a/a4/tcp/learnedModel.small.LStar.rand.dot b/a4/tcp/learnedModel.small.LStar.rand.dot
new file mode 100644
index 0000000..e169a68
--- /dev/null
+++ b/a4/tcp/learnedModel.small.LStar.rand.dot
@@ -0,0 +1,15 @@
+digraph g {
+__start0 [label="" shape="none"];
+
+	s0 [shape="circle" label="0"];
+	s1 [shape="circle" label="1"];
+	s2 [shape="circle" label="2"];
+	s0 -> s1 [label="SYN / SYN-ACK"];
+	s0 -> s0 [label="ACK / TO"];
+	s1 -> s1 [label="SYN / SYN-ACK"];
+	s1 -> s2 [label="ACK / TO"];
+	s2 -> s2 [label="SYN / TO"];
+	s2 -> s2 [label="ACK / TO"];
+
+__start0 -> s0;
+}
diff --git a/a4/tcp/run.sh b/a4/tcp/run.sh
index d3d70ef..645b4f8 100755
--- a/a4/tcp/run.sh
+++ b/a4/tcp/run.sh
@@ -1,2 +1,3 @@
 #!/bin/bash -x
-java -cp ":lib/automata-parent.jar:lib/learnlib-parent.jar" learner.Main
+#java -cp ":lib/automata-parent.jar:lib/learnlib-parent.jar" learner.Main small
+java -cp ":lib/automata-parent.jar:lib/learnlib-parent.jar" learner.Main partial LStar rand
diff --git a/a4/tcp/tester/learner/Main.java b/a4/tcp/tester/learner/Main.java
index 0ac2018..4d25e54 100644
--- a/a4/tcp/tester/learner/Main.java
+++ b/a4/tcp/tester/learner/Main.java
@@ -54,10 +54,7 @@ public class Main {
 	//*****************//
 	// Defines the input alphabet, adapt for your socket (you can even use other types than string, if you 
 	// change the generic-values, e.g. make your SUL of type SUL<Integer, Float> for int-input and float-output
-	private static final Alphabet<String> inputAlphabet =
-		new SimpleAlphabet<String>(ImmutableSet.of(
-					"SYN", "ACK",
-					"DAT", "RST", "FIN"));	
+	private static Alphabet<String> inputAlphabet;
 	// There are two SULs predefined, an example (see ExampleSul.java) and a socket SUL which connects to the SUL over socket
 	private static final SULType sulType = SULType.Socket;
 	public enum SULType { Example, Socket }
@@ -74,11 +71,11 @@ public class Main {
 	private static final String OUTPUT_FILENAME = "learnedModel";
 	// the learning and testing algorithms. LStar is the basic algorithm, TTT performs much faster
 	// but is a bit more inaccurate and produces more intermediate hypotheses, so test well)
-	private static final LearningMethod learningAlgorithm = LearningMethod.LStar;
+	private static LearningMethod learningAlgorithm = LearningMethod.LStar;
 	public enum LearningMethod { LStar, RivestSchapire, TTT, KearnsVazirani }
 	// Random walk is the simplest, but performs badly on large models: the chance of hitting a
 	// erroneous long trace is very small
-	private static final TestingMethod testMethod = TestingMethod.RandomWalk;
+	private static TestingMethod testMethod = TestingMethod.RandomWalk;
 	public enum TestingMethod { RandomWalk, WMethod, WpMethod }
 	// for random walk, the chance to do a reset after an input and the number of
 	// inputs to test before accepting a hypothesis
@@ -89,9 +86,80 @@ public class Main {
 	private static final boolean runControlledExperiment = true;
 	// For controlled experiments only: store every hypotheses as a file. Useful for 'debugging'
 	// if the learner does not terminate (hint: the TTT-algorithm produces many hypotheses).
-	private static final boolean saveAllHypotheses = false;
+	private static final boolean saveAllHypotheses = true;
+	
+	private static String stateCountStr;
+	private static String methodStr;
+	private static String testStr;
 	
 	public static void main(String [] args) throws IOException {
+		if (args.length != 3) {
+			System.out.println("error: missing commandline arguments!");
+			return;
+		}
+		
+		/* this option reduces the number of inputs used by the learning
+		 *
+		 * this should result in a smaller model because these inputs
+		 * are only valid in certain states.
+		 */
+		stateCountStr = args[0];
+		int stateCount = 0;
+		if (args[0].equals("partial")) {
+			stateCount = 1;
+		} else if (args[0].equals("full")) {
+			stateCount = 2;
+		}
+		
+		/* learning method
+		 *
+		 * can be LStar, TTT, RS or KV
+		 */
+		methodStr = args[1];
+		if (methodStr.equals("LStar")) {
+			learningAlgorithm = LearningMethod.LStar;
+		} else if (methodStr.equals("TTT")) {
+			learningAlgorithm = LearningMethod.TTT;
+		} else if (methodStr.equals("RS")) {
+			learningAlgorithm = LearningMethod.RivestSchapire;
+		} else if (methodStr.equals("KV")) {
+			learningAlgorithm = LearningMethod.KearnsVazirani;
+		}
+		
+		/* testing method
+		 *
+		 * can be: rand, wm, wpm
+		 */
+		testStr = args[2];
+		if (testStr.equals("rand")) {
+			testMethod = TestingMethod.RandomWalk;
+		} else if (testStr.equals("wm")) {
+			testMethod = TestingMethod.WMethod;
+		} else if (testStr.equals("wpm")) {
+			testMethod = TestingMethod.WpMethod;
+		}
+		
+		System.out.println("settings:");
+		System.out.println("stateCount = "+stateCount);
+		System.out.println("learningMethod = "+methodStr);
+		System.out.println("testMethod = "+testStr);
+		
+		switch (stateCount) {
+			case 2:
+				inputAlphabet = new SimpleAlphabet<String>(ImmutableSet.of(
+					"SYN", "ACK",
+					"DAT", "RST", "FIN"));
+			break;
+			case 1:
+				inputAlphabet = new SimpleAlphabet<String>(ImmutableSet.of(
+					"SYN", "ACK",
+					"RST"));
+			break;
+			default:
+				inputAlphabet = new SimpleAlphabet<String>(ImmutableSet.of(
+					"SYN", "ACK"));
+		}
+		
 		// Load the actual SUL-class, depending on which SUL-type is set at the top of this file
 		// You can also program an own SUL-class if you extend SUL<String,String> (or SUL<S,T> in
 		// general, with S and T the input and output types - you'll have to change some of the
@@ -258,9 +326,9 @@ public class Main {
 	 * @throws IOException
 	 */
 	public static void produceOutput(String fileName, MealyMachine<?,String,?,String> model, Alphabet<String> alphabet, boolean verboseError) throws FileNotFoundException, IOException {
-		GraphDOT.write(model, alphabet, new PrintWriter(OUTPUT_FILENAME + ".dot"));
+		GraphDOT.write(model, alphabet, new PrintWriter(fileName + "." + stateCountStr + "." + methodStr + "." + testStr + ".dot"));
 		try {
-			DOT.runDOT(new File(OUTPUT_FILENAME + ".dot"), "pdf", new File(OUTPUT_FILENAME + ".pdf"));
+			DOT.runDOT(new File(fileName + "." + stateCountStr + "." + methodStr + "." + testStr + ".dot"), "pdf", new File(fileName + "." + stateCountStr + "." + methodStr + "." + testStr + ".pdf"));
 		} catch (Exception e) {
 			if (verboseError) {
 				System.err.println("Warning: Install graphviz to convert dot-files to PDF");