review #1

javafx-framework/src/main/java/io/github/mzmine/javafx/components/factories/FxComponentFactory.java

@@ -27,18 +27,21 @@
 
 import javafx.beans.property.Property;
 import javafx.collections.ObservableList;
+import javafx.geometry.Pos;
 import javafx.scene.control.ComboBox;
 import javafx.scene.control.Label;
 import javafx.scene.layout.HBox;
 
 public class FxComponentFactory {
 
-  public static <T> HBox createLabelledComboBox(String label, ObservableList<T> values,
+  public static <T> HBox createLabeledComboBox(String label, ObservableList<T> values,
       Property<T> modelProperty) {
     final Label lab = new Label(label);
     final ComboBox<T> comboBox = new ComboBox<>(values);
     comboBox.valueProperty().bindBidirectional(modelProperty);
     lab.setLabelFor(comboBox);
-    return new HBox(5, lab, comboBox);
+    final HBox hBox = new HBox(5, lab, comboBox);
+    hBox.setAlignment(Pos.CENTER_LEFT);
+    return hBox;
   }
 }

mzmine-community/src/main/java/io/github/mzmine/datamodel/AbundanceMeasure.java

@@ -58,4 +58,11 @@ public Float get(@Nullable ModularDataModel featureOrRow) {
     return featureOrRow.get(type);
   }
 
+  public Float getOrNaN(@Nullable ModularDataModel featureOrRow) {
+    if (featureOrRow == null) {
+      return Float.NaN;
+    }
+    return featureOrRow.get(type);
+  }
+
 }

mzmine-community/src/main/java/io/github/mzmine/gui/chartbasics/chartthemes/EStandardChartTheme.java

@@ -76,6 +76,8 @@ public class EStandardChartTheme extends StandardChartTheme {
   private static final boolean DEFAULT_CROSS_HAIR_VISIBLE = true;
   private static final Stroke DEFAULT_CROSS_HAIR_STROKE = new BasicStroke(1.0F,
       BasicStroke.CAP_BUTT, BasicStroke.JOIN_BEVEL, 1.0f, new float[]{5.0F, 3.0F}, 0.0F);
+
+  public static final BasicStroke DEFAULT_STROKE = new BasicStroke(2f);
   // master font
   protected Font masterFont;
   protected Color masterFontColor;

mzmine-community/src/main/java/io/github/mzmine/gui/chartbasics/simplechart/renderers/ColoredXYShapeRenderer.java

@@ -62,9 +62,8 @@ public class ColoredXYShapeRenderer extends XYShapeRenderer {
 
   private static final int defaultSize = 7;
   private final Shape dataPointsShape;
-
   private final boolean drawOutlinesOnly;
-  private BasicStroke outlineStroke = new BasicStroke(2.0f);
+  private final BasicStroke outlineStroke = EStandardChartTheme.DEFAULT_STROKE;
 
   public ColoredXYShapeRenderer(boolean drawOutlinesOnly, Shape shape) {
     super();

mzmine-community/src/main/java/io/github/mzmine/gui/framework/fx/FxControllerBinding.java

@@ -29,8 +29,14 @@ public sealed interface FxControllerBinding permits SelectedAbundanceMeasureCont
     SelectedMetadataColumnController, SelectedRowsController, SelectedFeaturesController,
     SelectedFilesController, SelectedFeatureListsController {
 
+  public static void bindExposedProperties(Object master, Object child) {
+    if (master instanceof FxControllerBinding && child instanceof FxControllerBinding) {
+      bindExposedProperties(master, child);
+    }
+  }
+
   public static void bindExposedProperties(FxControllerBinding master, FxControllerBinding child) {
-    if(master == null ||child == null) {
+    if (master == null || child == null) {
       return;
     }
 

mzmine-community/src/main/java/io/github/mzmine/modules/dataanalysis/pca_new/PCAController.java

@@ -57,10 +57,6 @@ public ObjectProperty<List<FeatureListRow>> selectedRowsProperty() {
     return model.selectedRowsProperty();
   }
 
-  public ObjectProperty<List<FeatureList>> featureListsProperty() {
-    return model.flistsProperty();
-  }
-
   private void initListeners() {
     model.flistsProperty().addListener(_ -> update());
     model.domainPcProperty().addListener(_ -> update());

mzmine-community/src/main/java/io/github/mzmine/modules/dataanalysis/pca_new/LoadingsProvider.java → mzmine-community/src/main/java/io/github/mzmine/modules/dataanalysis/pca_new/PCALoadingsProvider.java

@@ -25,9 +25,6 @@
 
 package io.github.mzmine.modules.dataanalysis.pca_new;
 
-import static io.github.mzmine.util.annotations.CompoundAnnotationUtils.getAnnotationTypesByPriority;
-import static io.github.mzmine.util.annotations.CompoundAnnotationUtils.rankUniqueAnnotationTypes;
-
 import io.github.mzmine.datamodel.features.FeatureListRow;
 import io.github.mzmine.datamodel.features.types.DataType;
 import io.github.mzmine.datamodel.features.types.DataTypes;
@@ -38,6 +35,7 @@
 import io.github.mzmine.gui.chartbasics.simplechart.providers.ZCategoryProvider;
 import io.github.mzmine.main.MZmineCore;
 import io.github.mzmine.taskcontrol.TaskStatus;
+import io.github.mzmine.util.annotations.CompoundAnnotationUtils;
 import io.github.mzmine.util.color.SimpleColorPalette;
 import java.awt.Color;
 import java.util.Map;
@@ -47,33 +45,33 @@
 import org.jfree.chart.renderer.LookupPaintScale;
 import org.jfree.chart.renderer.PaintScale;
 
-public class LoadingsProvider extends SimpleXYProvider implements PlotXYZDataProvider,
+public class PCALoadingsProvider extends SimpleXYProvider implements PlotXYZDataProvider,
     ZCategoryProvider, XYItemObjectProvider<FeatureListRow> {
 
   private final PCARowsResult result;
-  private final int loadingsY;
-  private final int loadingsX;
+  private final int loadingsIndexY;
+  private final int loadingsIndexX;
 
   private int[] zCategories;
   private int numberOfCategories;
   private LookupPaintScale paintScale;
   private String[] legendNames;
 
   /**
-   * @param loadingsX index of the principal component used for domain axis, subtract 1 from the
-   *                  number since the pc matrix starts at 0.
-   * @param loadingsY index of the principal component used for range axis, subtract 1 from the
-   *                  number since the pc matrix starts at 0.
+   * @param loadingsIndexX index of the principal component used for domain axis, subtract 1 from
+   *                       the number since the pc matrix starts at 0.
+   * @param loadingsIndexY index of the principal component used for range axis, subtract 1 from the
+   *                       number since the pc matrix starts at 0.
    */
-  public LoadingsProvider(PCARowsResult result, String seriesKey, Color awt, int loadingsX,
-      int loadingsY) {
+  public PCALoadingsProvider(PCARowsResult result, String seriesKey, Color awt, int loadingsIndexX,
+      int loadingsIndexY) {
     super(seriesKey, awt);
     this.result = result;
-    this.loadingsX = loadingsX;
-    this.loadingsY = loadingsY;
+    this.loadingsIndexX = loadingsIndexX;
+    this.loadingsIndexY = loadingsIndexY;
   }
 
-  public LoadingsProvider(PCARowsResult result, String seriesKey, Color awt) {
+  public PCALoadingsProvider(PCARowsResult result, String seriesKey, Color awt) {
     this(result, seriesKey, awt, 0, 1);
   }
 
@@ -83,11 +81,12 @@ public void computeValues(Property<TaskStatus> status) {
 
     final RealMatrix loadingsMatrix = pcaResult.getLoadingsMatrix();
 
-    final Map<FeatureListRow, DataType<?>> bestRowAnnotationType = getAnnotationTypesByPriority(
+    final Map<FeatureListRow, DataType<?>> bestRowAnnotationType = CompoundAnnotationUtils.mapBestAnnotationTypesByPriority(
         result.rows(), true);
 
     // only create order of actually existing annotaiton types
-    Map<DataType<?>, Integer> typesInOrder = rankUniqueAnnotationTypes(bestRowAnnotationType.values());
+    Map<DataType<?>, Integer> typesInOrder = CompoundAnnotationUtils.rankUniqueAnnotationTypes(
+        bestRowAnnotationType.values());
     numberOfCategories = typesInOrder.size();
 
     double[] domainData = new double[loadingsMatrix.getColumnDimension()];
@@ -97,8 +96,8 @@ public void computeValues(Property<TaskStatus> status) {
 
     final MissingValueType missing = DataTypes.get(MissingValueType.class);
     for (int i = 0; i < loadingsMatrix.getColumnDimension(); i++) {
-      domainData[i] = loadingsMatrix.getEntry(loadingsX, i);
-      rangeData[i] = loadingsMatrix.getEntry(loadingsY, i);
+      domainData[i] = loadingsMatrix.getEntry(loadingsIndexX, i);
+      rangeData[i] = loadingsMatrix.getEntry(loadingsIndexY, i);
       // find annotation type or missing type
       FeatureListRow row = result.rows().get(i);
       final DataType<?> bestTypeWithValue = bestRowAnnotationType.get(row);

mzmine-community/src/main/java/io/github/mzmine/modules/dataanalysis/pca_new/PCAResult.java

@@ -31,18 +31,36 @@
 import org.apache.commons.math3.linear.SingularValueDecomposition;
 import org.jetbrains.annotations.NotNull;
 
+/**
+ * A pca based on singular value decomposition. The data matrix X is decomposed into X = U*S*V.
+ * Columns of U contrain the principal components, S are the singular values, which can be projected
+ * into the PC space using U and a submatrix of S, which creates the scores plot. Loadings are the
+ * transpose of V.
+ *
+ * https://stats.stackexchange.com/questions/134282/relationship-between-svd-and-pca-how-to-use-svd-to-perform-pca
+ */
 public record PCAResult(RealMatrix data, RealMatrix dataMeanCentered,
                         SingularValueDecomposition svd) {
 
+  /**
+   * @param numComponents
+   * @return Returns a sub-matrix the first n principal components of the decomposition.
+   */
   public RealMatrix firstNComponents(int numComponents) {
-    return svd.getU().getSubMatrix(0, svd.getU().getRowDimension() - 1, 0,
-        numComponents - 1);
+    return svd.getU().getSubMatrix(0, svd.getU().getRowDimension() - 1, 0, numComponents - 1);
   }
 
   public RealMatrix principalComponentsMatrix() {
+    // the u matrix of an svd contains the principal components.
     return svd.getU();
   }
 
+  /**
+   * Projects the data matrix onto the principal components. The result is n dimensional and is
+   * called "scores" matrix and is used for the scores plot.
+   *
+   * @param numComponents the number of components n.
+   */
   public RealMatrix projectDataToScores(int numComponents) {
     final RealMatrix firstNComponents = firstNComponents(numComponents);
     final RealMatrix subMatrixS = svd.getS()
@@ -51,15 +69,25 @@ public RealMatrix projectDataToScores(int numComponents) {
     return projectedData;
   }
 
+  /**
+   * Projects the data matrix onto the selected principal components. The result is 2 dimensional
+   * and is called "scores" matrix and is used for the scores plot.
+   *
+   * @see #projectDataToScores(int)
+   */
   public RealMatrix projectDataToScores(int domainColIndex, int rangeColIndex) {
     final RealMatrix pcMatrix = pcMatrix(domainColIndex, rangeColIndex);
     final RealMatrix projected = pcMatrix.multiply(svd.getS().getSubMatrix(0, 1, 0, 1));
     return projected;
   }
 
+  /**
+   * Retrieves two specific PCs from the PC matrix.
+   */
   @NotNull
   private RealMatrix pcMatrix(int domainColIndex, int rangeColIndex) {
     final RealMatrix pcs = svd.getU();
+    // the vectors are the respective components.
     final RealVector domainVector = pcs.getColumnVector(domainColIndex);
     final RealVector rangeVector = pcs.getColumnVector(rangeColIndex);
     RealMatrix pcMatrix = new Array2DRowRealMatrix(pcs.getRowDimension(), 2);
@@ -69,8 +97,16 @@ private RealMatrix pcMatrix(int domainColIndex, int rangeColIndex) {
     return pcMatrix;
   }
 
+  /**
+   * Retrieves the loadings (importance of each observed feature) from the pca. After svd the
+   * loadings are the transpose of the v matrix.
+   */
   public RealMatrix getLoadingsMatrix() {
     final RealMatrix transpose = svd.getV().transpose();
     return transpose;
   }
+
+  public int componentCount() {
+    return principalComponentsMatrix().getRowDimension();
+  }
 }

mzmine-community/src/main/java/io/github/mzmine/modules/dataanalysis/pca_new/ScoresProvider.java → mzmine-community/src/main/java/io/github/mzmine/modules/dataanalysis/pca_new/PCAScoresProvider.java

@@ -47,7 +47,7 @@
 import org.jfree.chart.renderer.LookupPaintScale;
 import org.jfree.chart.renderer.PaintScale;
 
-public class ScoresProvider extends SimpleXYProvider implements PlotXYZDataProvider,
+public class PCAScoresProvider extends SimpleXYProvider implements PlotXYZDataProvider,
     ZCategoryProvider, XYItemObjectProvider<RawDataFile> {
 
   private final PCARowsResult result;
@@ -66,7 +66,7 @@ public class ScoresProvider extends SimpleXYProvider implements PlotXYZDataProvi
    * @param pcY index of the principal component used for range axis, subtract 1 from the number
    *            since the pc matrix starts at 0.
    */
-  public ScoresProvider(PCARowsResult result, String seriesKey, Color awt, int pcX, int pcY,
+  public PCAScoresProvider(PCARowsResult result, String seriesKey, Color awt, int pcX, int pcY,
       MetadataColumn<?> groupingColumn) {
     super(seriesKey, awt);
     this.result = result;
@@ -75,7 +75,7 @@ public ScoresProvider(PCARowsResult result, String seriesKey, Color awt, int pcX
     this.groupingColumn = groupingColumn;
   }
 
-  public ScoresProvider(PCARowsResult result, String seriesKey, Color awt) {
+  public PCAScoresProvider(PCARowsResult result, String seriesKey, Color awt) {
     this(result, seriesKey, awt, 0, 1, null);
   }
 
@@ -86,9 +86,8 @@ public void computeValues(Property<TaskStatus> status) {
     final RealMatrix scores = pcaResult.projectDataToScores(pcX, pcY);
 
     final List<RawDataFile> files = result.files();
-    final Map<?, List<RawDataFile>> groupedFiles =
-        groupingColumn != null ? MZmineCore.getProjectMetadata().groupFilesByColumn(groupingColumn)
-            : Map.of();
+    final Map<?, List<RawDataFile>> groupedFiles = MZmineCore.getProjectMetadata()
+        .groupFilesByColumn(groupingColumn);
 
     numberOfCategories = Math.max(groupedFiles.size(), 1);
     groupNames = new String[Math.max(groupedFiles.size(), 1)];

mzmine-community/src/main/java/io/github/mzmine/modules/dataanalysis/pca_new/PCAUpdateTask.java

@@ -93,13 +93,13 @@ protected void process() {
     pcaRowsResult = PCAUtils.performPCAOnRows(flists.get(0).getRows(), abundance);
     progressProvider.getAndIncrement();
 
-    final ScoresProvider scores = new ScoresProvider(pcaRowsResult, "Scores", Color.RED,
+    final PCAScoresProvider scores = new PCAScoresProvider(pcaRowsResult, "Scores", Color.RED,
         domainPcIndex, rangePcIndex,
         MZmineCore.getProjectMetadata().getColumnByName(metadataColumn));
     final ColoredXYZDataset scoresDS = new ColoredXYZDataset(scores, RunOption.THIS_THREAD);
     progressProvider.getAndIncrement();
 
-    final LoadingsProvider loadings = new LoadingsProvider(pcaRowsResult, "Loadings", Color.RED,
+    final PCALoadingsProvider loadings = new PCALoadingsProvider(pcaRowsResult, "Loadings", Color.RED,
         domainPcIndex, rangePcIndex);
     final ColoredXYZDataset loadingsDS = new ColoredXYZDataset(loadings, RunOption.THIS_THREAD);
     progressProvider.getAndIncrement();
@@ -123,12 +123,12 @@ protected void updateGuiModel() {
     if (rangePcIndex < components.size()) {
       model.setRangePc(rangePcIndex + 1);
     } else {
-      model.setRangePc(components.getLast() + 1);
+      model.setRangePc(components.getLast());
     }
     if (domainPcIndex < components.size()) {
       model.setDomainPc(domainPcIndex + 1);
     } else {
-      model.setDomainPc(components.getLast() + 1);
+      model.setDomainPc(components.getLast());
     }
   }
 

mzmine-community/src/main/java/io/github/mzmine/modules/dataanalysis/pca_new/PCAUtils.java

@@ -38,6 +38,13 @@ public class PCAUtils {
 
   private static final Logger logger = Logger.getLogger(PCAUtils.class.getName());
 
+  /**
+   * Calculates the PCA of a matrix by singular value decomposition (svd).
+   * https://stats.stackexchange.com/questions/134282/relationship-between-svd-and-pca-how-to-use-svd-to-perform-pca
+   *
+   * @param data the data
+   * @return A pca result.
+   */
   public static PCAResult calculatePCA(RealMatrix data) {
 
     logger.finest(() -> "Performing scaling mean centering");
@@ -49,6 +56,13 @@ public static PCAResult calculatePCA(RealMatrix data) {
     return new PCAResult(data, centeredMatrix, svd);
   }
 
+  /**
+   * Performs a PCA on a list of feature list rows. Imputes missing values as 0s.
+   *
+   * @param rows    The rows.
+   * @param measure The abundance to use.
+   * @return A pca result that can be mapped to the used rows.
+   */
   public static PCARowsResult performPCAOnRows(List<FeatureListRow> rows,
       AbundanceMeasure measure) {
     final List<RawDataFile> files = rows.stream().flatMap(row -> row.getRawDataFiles().stream())