X-Git-Url: http://source.jalview.org/gitweb/?a=blobdiff_plain;f=src%2Fjalview%2Fviewmodel%2FPCAModel.java;h=928d35e46ccc5199bf76dfa03a5506c2f9d48d62;hb=c1f777a6a4a74ee1573e5f3f883e0de902818ae7;hp=30f87837ef26a41fab717122bf64efb63e1abcdc;hpb=55c883270e9f64da9562a24f09dfe5f2f079e59a;p=jalview.git
diff --git a/src/jalview/viewmodel/PCAModel.java b/src/jalview/viewmodel/PCAModel.java
index 30f8783..928d35e 100644
--- a/src/jalview/viewmodel/PCAModel.java
+++ b/src/jalview/viewmodel/PCAModel.java
@@ -1,43 +1,80 @@
+/*
+ * Jalview - A Sequence Alignment Editor and Viewer ($$Version-Rel$$)
+ * Copyright (C) $$Year-Rel$$ The Jalview Authors
+ *
+ * This file is part of Jalview.
+ *
+ * Jalview is free software: you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License
+ * as published by the Free Software Foundation, either version 3
+ * of the License, or (at your option) any later version.
+ *
+ * Jalview is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty
+ * of MERCHANTABILITY or FITNESS FOR A PARTICULAR
+ * PURPOSE. See the GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with Jalview. If not, see .
+ * The Jalview Authors are detailed in the 'AUTHORS' file.
+ */
package jalview.viewmodel;
-import java.util.Vector;
-
import jalview.analysis.PCA;
+import jalview.api.RotatableCanvasI;
+import jalview.api.analysis.ScoreModelI;
+import jalview.api.analysis.SimilarityParamsI;
import jalview.datamodel.AlignmentView;
import jalview.datamodel.SequenceI;
import jalview.datamodel.SequencePoint;
-import jalview.api.RotatableCanvasI;
+
+import java.util.Vector;
public class PCAModel
{
+ private volatile PCA pca;
- public PCAModel(AlignmentView seqstrings2, SequenceI[] seqs2,
- boolean nucleotide2)
- {
- seqstrings=seqstrings2;
- seqs=seqs2;
- nucleotide=nucleotide2;
- }
-
- PCA pca;
-
int top;
-
+
AlignmentView seqstrings;
SequenceI[] seqs;
- /**
- * use the identity matrix for calculating similarity between sequences.
+ /*
+ * Name of score model used to calculate PCA
*/
- private boolean nucleotide=false;
+ ScoreModelI scoreModel;
+
+ private boolean nucleotide = false;
private Vector points;
+ private SimilarityParamsI similarityParams;
+
+ /**
+ * Constructor given sequence data, score model and score calculation
+ * parameter options.
+ *
+ * @param seqData
+ * @param sqs
+ * @param nuc
+ * @param modelName
+ * @param params
+ */
+ public PCAModel(AlignmentView seqData, SequenceI[] sqs, boolean nuc,
+ ScoreModelI modelName,
+ SimilarityParamsI params)
+ {
+ seqstrings = seqData;
+ seqs = sqs;
+ nucleotide = nuc;
+ scoreModel = modelName;
+ similarityParams = params;
+ }
+
public void run()
{
-
- pca = new PCA(seqstrings.getSequenceStrings(' '), nucleotide);
+ pca = new PCA(seqstrings, scoreModel, similarityParams);
pca.run();
// Now find the component coordinates
@@ -48,41 +85,33 @@ public class PCAModel
ii++;
}
- double[][] comps = new double[ii][ii];
-
- for (int i = 0; i < ii; i++)
- {
- if (pca.getEigenvalue(i) > 1e-4)
- {
- comps[i] = pca.component(i);
- }
- }
-
- top = pca.getM().rows - 1;
+ int height = pca.getHeight();
+ // top = pca.getM().height() - 1;
+ top = height - 1;
points = new Vector();
float[][] scores = pca.getComponents(top - 1, top - 2, top - 3, 100);
- for (int i = 0; i < pca.getM().rows; i++)
+ for (int i = 0; i < height; i++)
{
SequencePoint sp = new SequencePoint(seqs[i], scores[i]);
points.addElement(sp);
}
-
}
public void updateRc(RotatableCanvasI rc)
{
- rc.setPoints(points, pca.getM().rows);
+ rc.setPoints(points, pca.getHeight());
}
public boolean isNucleotide()
{
return nucleotide;
}
+
public void setNucleotide(boolean nucleotide)
{
- this.nucleotide=nucleotide;
+ this.nucleotide = nucleotide;
}
/**
@@ -97,19 +126,22 @@ public class PCAModel
/**
* update the 2d coordinates for the list of points to the given dimensions
- * Principal dimension is getTop(). Next greated eigenvector is getTop()-1.
- * Note - pca.getComponents starts counting the spectrum from zero rather than one, so getComponents(dimN ...) == updateRcView(dimN+1 ..)
- * @param dim1
+ * Principal dimension is getTop(). Next greatest eigenvector is getTop()-1.
+ * Note - pca.getComponents starts counting the spectrum from rank-2 to zero,
+ * rather than rank-1, so getComponents(dimN ...) == updateRcView(dimN+1 ..)
+ *
+ * @param dim1
* @param dim2
* @param dim3
*/
public void updateRcView(int dim1, int dim2, int dim3)
{
- float[][] scores = pca.getComponents(dim1-1, dim2-1, dim3-1, 100);
+ // note: actual indices for components are dim1-1, etc (patch for JAL-1123)
+ float[][] scores = pca.getComponents(dim1 - 1, dim2 - 1, dim3 - 1, 100);
- for (int i = 0; i < pca.getM().rows; i++)
+ for (int i = 0; i < pca.getHeight(); i++)
{
- ((SequencePoint) points.elementAt(i)).coord = scores[i];
+ points.elementAt(i).coord = scores[i];
}
}
@@ -122,7 +154,9 @@ public class PCAModel
{
return seqstrings;
}
- public String getPointsasCsv(boolean transformed, int xdim, int ydim, int zdim)
+
+ public String getPointsasCsv(boolean transformed, int xdim, int ydim,
+ int zdim)
{
StringBuffer csv = new StringBuffer();
csv.append("\"Sequence\"");
@@ -187,4 +221,14 @@ public class PCAModel
return pts;
}
+ public String getScoreModelName()
+ {
+ return scoreModel == null ? "" : scoreModel.getName();
+ }
+
+ public void setScoreModel(ScoreModelI sm)
+ {
+ this.scoreModel = sm;
+ }
+
}