WIP: don't merge me

.github/workflows/capymoa.yml

@@ -0,0 +1,50 @@
+name: CapyMOA Test and Package
+
+on:
+  push:
+    branches: [ master ]
+  pull_request:
+    branches: [ master ]
+
+jobs:
+  build:
+
+    runs-on: ubuntu-latest
+
+    steps:
+    - uses: actions/checkout@v4
+
+    # TODO: moa tests are sensitive to java versions https://github.com/Waikato/moa/issues/273
+    - name: Set up JDK 21
+      uses: actions/setup-java@v4
+      with:
+        java-version: '21'
+        distribution: 'zulu'
+        cache: maven
+
+    - name: Version
+      working-directory: ./moa
+      run: mvn -v
+
+    - name: Unit Tests
+      working-directory: ./moa
+      # -B: non-interactive (batch) mode
+      # -q: quiet output
+      run: mvn -B -q test 
+
+    - name: Package Jar
+      working-directory: ./moa
+      # -B: non-interactive (batch) mode
+      # -DskipTests: skip tests (they were run earlier)
+      # -Dmaven.javadoc.skip=true: skip javadoc generation
+      # -Dlatex.skipBuild=true: skip latex documentation build this needs extra dependencies
+      run: mvn -B package -DskipTests -Dmaven.javadoc.skip=true -Dlatex.skipBuild=true
+
+    # Upload jar file as artifact
+    - name: Upload Jar
+      uses: actions/upload-artifact@v4
+      with:
+        name: moa-jar
+        path: ./moa/target/moa-*-jar-with-dependencies.jar
+        if-no-files-found: error
+        retention-days: 7

.gitignore

@@ -3,4 +3,31 @@
 *.iml
 *~
 *.bak
+.DS_Store
+.settings
+.project
 
+# Compiled class file
+*.class
+
+# Log file
+*.log
+
+# BlueJ files
+*.ctxt
+
+# Mobile Tools for Java (J2ME)
+.mtj.tmp/
+
+# Package Files #
+*.jar
+*.war
+*.nar
+*.ear
+*.zip
+*.tar.gz
+*.rar
+
+# virtual machine crash logs, see http://www.java.com/en/download/help/error_hotspot.xml
+hs_err_pid*
+replay_pid*

CapyMOA.md

@@ -0,0 +1,29 @@
+# CapyMOA
+
+CapyMOA is a datastream learning framework that integrates the [Massive Online Analysis
+(MOA)](https://moa.cms.waikato.ac.nz/) library with the python ecosystem.
+
+To build MOA for use with CapyMOA run:
+```bash
+cd moa
+mvn package -DskipTests -Dmaven.javadoc.skip=true -Dlatex.skipBuild=true
+```
+This will create a `target/moa-*-jar-with-dependencies.jar` file that can be used by
+CapyMOA. To let CapyMOA know where this file is, set the `CAPYMOA_MOA_JAR` environment
+variable to the path of this file.
+
+You can do this temporarily in your terminal session with:
+```bash
+export CAPYMOA_MOA_JAR=/path/to/moa/target/moa-*-jar-with-dependencies.jar
+```
+To check that CapyMOA can find MOA, run:
+```bash
+python -c "import capymoa; capymoa.about()"
+# CapyMOA 0.10.0
+#   CAPYMOA_DATASETS_DIR: .../datasets
+#   CAPYMOA_MOA_JAR:      .../moa/moa/target/moa-2024.07.2-SNAPSHOT-jar-with-dependencies.jar
+#   CAPYMOA_JVM_ARGS:     ['-Xmx8g', '-Xss10M']
+#   JAVA_HOME:            /usr/lib/jvm/java-21-openjdk
+#   MOA version:          aa955ebbcbd99e9e1d19ab16582e3e5a6fca5801ba250e4d164c16a89cf798ea
+#   JAVA version:         21.0.7
+```

README.md

@@ -30,5 +30,3 @@ If you want to refer to MOA in a publication, please cite the following JMLR pap
 
 > Albert Bifet, Geoff Holmes, Richard Kirkby, Bernhard Pfahringer (2010);
 > MOA: Massive Online Analysis; Journal of Machine Learning Research 11: 1601-1604 
-
-

moa/pom.xml

@@ -259,6 +259,16 @@
       <plugin>
         <groupId>org.apache.maven.plugins</groupId>
         <artifactId>maven-assembly-plugin</artifactId>
+        <configuration>
+          <archive>
+            <manifest>
+              <mainClass>moa.gui.GUI</mainClass>
+            </manifest>
+          </archive>
+          <descriptorRefs>
+            <descriptorRef>jar-with-dependencies</descriptorRef>
+          </descriptorRefs>
+        </configuration>
       </plugin>
 
       <plugin>

moa/src/main/java/moa/classifiers/AbstractClassifier.java

@@ -105,6 +105,26 @@ public double[] getVotesForInstance(Example<Instance> example){
     @Override
     public abstract double[] getVotesForInstance(Instance inst);
 
+    @Override
+    public double getConfidenceForPrediction(Instance inst, double prediction) {
+        double[] votes = this.getVotesForInstance(inst);
+        double predictionValue = votes[(int) prediction];
+
+        double sum = 0.0;
+        for (double vote : votes)
+            sum += vote;
+
+        // Check if the sum is zero
+        if (sum == 0.0)
+            return 0.0; // Return 0 if sum is zero to avoid division by zero
+        return predictionValue / sum;
+    }
+
+    @Override
+    public double getConfidenceForPrediction(Example<Instance> example, double prediction) {
+        return getConfidenceForPrediction(example.getData(), prediction);
+    }
+
     @Override
     public Prediction getPredictionForInstance(Example<Instance> example){
 		return getPredictionForInstance(example.getData());

moa/src/main/java/moa/classifiers/Classifier.java

@@ -15,7 +15,7 @@
  *
  *    You should have received a copy of the GNU General Public License
  *    along with this program. If not, see <http://www.gnu.org/licenses/>.
- *    
+ *
  */
 package moa.classifiers;
 
@@ -76,7 +76,7 @@ public interface Classifier extends Learner<Example<Instance>> {
      * test instance in each class
      */
     public double[] getVotesForInstance(Instance inst);
-    
+
     /**
      * Sets the reference to the header of the data stream. The header of the
      * data stream is extended from WEKA
@@ -86,7 +86,7 @@ public interface Classifier extends Learner<Example<Instance>> {
      * @param ih the reference to the data stream header
      */
     //public void setModelContext(InstancesHeader ih);
-    
+
     /**
      * Gets the reference to the header of the data stream. The header of the
      * data stream is extended from WEKA
@@ -96,6 +96,8 @@ public interface Classifier extends Learner<Example<Instance>> {
      * @return the reference to the data stream header
      */
     //public InstancesHeader getModelContext();
-    
+
     public Prediction getPredictionForInstance(Instance inst);
+
+    public double getConfidenceForPrediction(Instance inst, double prediction);
 }

moa/src/main/java/moa/classifiers/SemiSupervisedLearner.java

@@ -19,12 +19,16 @@
  */
 package moa.classifiers;
 
+import com.yahoo.labs.samoa.instances.Instance;
+import moa.core.Example;
+import moa.learners.Learner;
+
 /**
- * Learner interface for incremental semi supervised models. It is used only in the GUI Regression Tab. 
- *
- * @author Richard Kirkby (rkirkby@cs.waikato.ac.nz)
- * @version $Revision: 7 $
+ * Updated learner interface for semi-supervised methods.
  */
-public interface SemiSupervisedLearner {
-
+public interface SemiSupervisedLearner extends Learner<Example<Instance>> {
+    // Returns the pseudo-label used. If no pseudo-label was used, then return -1.
+    int trainOnUnlabeledInstance(Instance instance);
+
+    void addInitialWarmupTrainingInstances();
 }