-/*\r
-* Jalview - A Sequence Alignment Editor and Viewer\r
-* Copyright (C) 2005 AM Waterhouse, J Procter, G Barton, M Clamp, S Searle\r
-*\r
-* This program is free software; you can redistribute it and/or\r
-* modify it under the terms of the GNU General Public License\r
-* as published by the Free Software Foundation; either version 2\r
-* of the License, or (at your option) any later version.\r
-*\r
-* This program is distributed in the hope that it will be useful,\r
-* but WITHOUT ANY WARRANTY; without even the implied warranty of\r
-* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the\r
-* GNU General Public License for more details.\r
-*\r
-* You should have received a copy of the GNU General Public License\r
-* along with this program; if not, write to the Free Software\r
-* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA\r
-*/\r
-package jalview.analysis;\r
-\r
-import jalview.datamodel.*;\r
-\r
-import jalview.math.*;\r
-\r
-import jalview.util.*;\r
-\r
-import java.awt.*;\r
-\r
-import java.io.*;\r
-\r
-\r
-public class PCA implements Runnable {\r
- Matrix m;\r
- Matrix symm;\r
- Matrix m2;\r
- double[] eigenvalue;\r
- Matrix eigenvector;\r
-\r
- public PCA(Matrix m) {\r
- this.m = m;\r
- }\r
-\r
- public PCA(SequenceI[] s) {\r
- Runtime rt = Runtime.getRuntime();\r
-\r
- BinarySequence[] bs = new BinarySequence[s.length];\r
- int ii = 0;\r
-\r
- while ((ii < s.length) && (s[ii] != null)) {\r
- bs[ii] = new BinarySequence(s[ii]);\r
- bs[ii].encode();\r
- ii++;\r
- }\r
-\r
- BinarySequence[] bs2 = new BinarySequence[s.length];\r
- ii = 0;\r
-\r
- while ((ii < s.length) && (s[ii] != null)) {\r
- bs2[ii] = new BinarySequence(s[ii]);\r
- bs2[ii].blosumEncode();\r
- ii++;\r
- }\r
-\r
- //System.out.println("Created binary encoding");\r
- //printMemory(rt);\r
- int count = 0;\r
-\r
- while ((count < bs.length) && (bs[count] != null)) {\r
- count++;\r
- }\r
-\r
- double[][] seqmat = new double[count][bs[0].getDBinary().length];\r
- double[][] seqmat2 = new double[count][bs2[0].getDBinary().length];\r
- int i = 0;\r
-\r
- while (i < count) {\r
- seqmat[i] = bs[i].getDBinary();\r
- seqmat2[i] = bs2[i].getDBinary();\r
- i++;\r
- }\r
-\r
- //System.out.println("Created array");\r
- //printMemory(rt);\r
- // System.out.println(" --- Original matrix ---- ");\r
- m = new Matrix(seqmat, count, bs[0].getDBinary().length);\r
- m2 = new Matrix(seqmat2, count, bs2[0].getDBinary().length);\r
-\r
- //System.out.println("Created matrix");\r
- printMemory(rt);\r
- }\r
-\r
- public static void printMemory(Runtime rt) {\r
- System.out.println("PCA:Free memory = " + rt.freeMemory());\r
- }\r
-\r
- public Matrix getM() {\r
- return m;\r
- }\r
-\r
- public double[] getEigenvector(int i) {\r
- return eigenvector.getColumn(i);\r
- }\r
-\r
- public double getEigenvalue(int i) {\r
- return eigenvector.d[i];\r
- }\r
-\r
- public float[][] getComponents(int l, int n, int mm) {\r
- return getComponents(l, n, mm, 1);\r
- }\r
-\r
- public float[][] getComponents(int l, int n, int mm, float factor) {\r
- float[][] out = new float[m.rows][3];\r
-\r
- for (int i = 0; i < m.rows; i++) {\r
- out[i][0] = (float) component(i, l) * factor;\r
- out[i][1] = (float) component(i, n) * factor;\r
- out[i][2] = (float) component(i, mm) * factor;\r
- }\r
-\r
- return out;\r
- }\r
-\r
- public double[] component(int n) {\r
- // n = index of eigenvector\r
- double[] out = new double[m.rows];\r
-\r
- for (int i = 0; i < m.rows; i++) {\r
- out[i] = component(i, n);\r
- }\r
-\r
- return out;\r
- }\r
-\r
- public double component(int row, int n) {\r
- double out = 0.0;\r
-\r
- for (int i = 0; i < symm.cols; i++) {\r
- out += (symm.value[row][i] * eigenvector.value[i][n]);\r
- }\r
-\r
- return out / eigenvector.d[n];\r
- }\r
-\r
- public void checkEigenvector(int n, PrintStream ps) {\r
- ps.println(" --- Eigenvector " + n + " --- ");\r
-\r
- double[] eigenv = eigenvector.getColumn(n);\r
-\r
- for (int i = 0; i < eigenv.length; i++) {\r
- Format.print(ps, "%15.4f", eigenv[i]);\r
- }\r
-\r
- System.out.println();\r
-\r
- double[] neigenv = symm.vectorPostMultiply(eigenv);\r
- System.out.println(" --- symmat * eigenv / lambda --- ");\r
-\r
- if (eigenvector.d[n] > 1e-4) {\r
- for (int i = 0; i < neigenv.length; i++) {\r
- Format.print(System.out, "%15.4f", neigenv[i] / eigenvector.d[n]);\r
- }\r
- }\r
-\r
- System.out.println();\r
- }\r
-\r
- public void run() {\r
- Matrix mt = m.transpose();\r
-\r
- // System.out.println(" --- OrigT * Orig ---- ");\r
- eigenvector = mt.preMultiply(m2);\r
-\r
- // eigenvector.print(System.out);\r
- symm = eigenvector.copy();\r
-\r
- //TextArea ta = new TextArea(25,72);\r
- //TextAreaPrintStream taps = new TextAreaPrintStream(System.out,ta);\r
- //Frame f = new Frame("PCA output");\r
- //f.resize(500,500);\r
- //f.setLayout(new BorderLayout());\r
- //f.add("Center",ta);\r
- //f.show();\r
- //symm.print(taps);\r
- long tstart = System.currentTimeMillis();\r
- eigenvector.tred();\r
-\r
- long tend = System.currentTimeMillis();\r
-\r
- //taps.println("Time take for tred = " + (tend-tstart) + "ms");\r
- //taps.println(" ---Tridiag transform matrix ---");\r
- //taps.println(" --- D vector ---");\r
- //eigenvector.printD(taps);\r
- //taps.println();\r
- //taps.println(" --- E vector ---");\r
- // eigenvector.printE(taps);\r
- //taps.println();\r
- // Now produce the diagonalization matrix\r
- tstart = System.currentTimeMillis();\r
- eigenvector.tqli();\r
- tend = System.currentTimeMillis();\r
-\r
- //System.out.println("Time take for tqli = " + (tend-tstart) + " ms");\r
- //System.out.println(" --- New diagonalization matrix ---");\r
- //System.out.println(" --- Eigenvalues ---");\r
- //eigenvector.printD(taps);\r
- //System.out.println();\r
- // for (int i=0; i < eigenvector.cols; i++) {\r
- // checkEigenvector(i,taps);\r
- // taps.println();\r
- // }\r
- // taps.println();\r
- // taps.println("Transformed sequences = ");\r
- // Matrix trans = m.preMultiply(eigenvector);\r
- // trans.print(System.out);\r
- }\r
-}\r
+/*
+ * 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 <http://www.gnu.org/licenses/>.
+ * The Jalview Authors are detailed in the 'AUTHORS' file.
+ */
+package jalview.analysis;
+
+import jalview.api.analysis.ScoreModelI;
+import jalview.api.analysis.SimilarityParamsI;
+import jalview.datamodel.AlignmentView;
+import jalview.math.MatrixI;
+
+import java.io.PrintStream;
+
+/**
+ * Performs Principal Component Analysis on given sequences
+ */
+public class PCA implements Runnable
+{
+ /*
+ * inputs
+ */
+ final private AlignmentView seqs;
+
+ final private ScoreModelI scoreModel;
+
+ final private SimilarityParamsI similarityParams;
+
+ /*
+ * outputs
+ */
+ private MatrixI symm;
+
+ private MatrixI eigenvector;
+
+ private String details;
+
+ /**
+ * Constructor given the sequences to compute for, the similarity model to
+ * use, and a set of parameters for sequence comparison
+ *
+ * @param sequences
+ * @param sm
+ * @param options
+ */
+ public PCA(AlignmentView sequences, ScoreModelI sm, SimilarityParamsI options)
+ {
+ this.seqs = sequences;
+ this.scoreModel = sm;
+ this.similarityParams = options;
+ }
+
+ /**
+ * Returns Eigenvalue
+ *
+ * @param i
+ * Index of diagonal within matrix
+ *
+ * @return Returns value of diagonal from matrix
+ */
+ public double getEigenvalue(int i)
+ {
+ return eigenvector.getD()[i];
+ }
+
+ /**
+ * DOCUMENT ME!
+ *
+ * @param l
+ * DOCUMENT ME!
+ * @param n
+ * DOCUMENT ME!
+ * @param mm
+ * DOCUMENT ME!
+ * @param factor
+ * DOCUMENT ME!
+ *
+ * @return DOCUMENT ME!
+ */
+ public float[][] getComponents(int l, int n, int mm, float factor)
+ {
+ float[][] out = new float[getHeight()][3];
+
+ for (int i = 0; i < getHeight(); i++)
+ {
+ out[i][0] = (float) component(i, l) * factor;
+ out[i][1] = (float) component(i, n) * factor;
+ out[i][2] = (float) component(i, mm) * factor;
+ }
+
+ return out;
+ }
+
+ /**
+ * DOCUMENT ME!
+ *
+ * @param n
+ * DOCUMENT ME!
+ *
+ * @return DOCUMENT ME!
+ */
+ public double[] component(int n)
+ {
+ // n = index of eigenvector
+ double[] out = new double[getHeight()];
+
+ for (int i = 0; i < out.length; i++)
+ {
+ out[i] = component(i, n);
+ }
+
+ return out;
+ }
+
+ /**
+ * DOCUMENT ME!
+ *
+ * @param row
+ * DOCUMENT ME!
+ * @param n
+ * DOCUMENT ME!
+ *
+ * @return DOCUMENT ME!
+ */
+ double component(int row, int n)
+ {
+ double out = 0.0;
+
+ for (int i = 0; i < symm.width(); i++)
+ {
+ out += (symm.getValue(row, i) * eigenvector.getValue(i, n));
+ }
+
+ return out / eigenvector.getD()[n];
+ }
+
+ /**
+ * Answers a formatted text report of the PCA calculation results (matrices
+ * and eigenvalues) suitable for display
+ *
+ * @return
+ */
+ public String getDetails()
+ {
+ return details;
+ }
+
+ /**
+ * Performs the PCA calculation
+ */
+ @Override
+ public void run()
+ {
+ /*
+ * print details to a string buffer as they are computed
+ */
+ StringBuilder sb = new StringBuilder(1024);
+ sb.append("PCA calculation using ").append(scoreModel.getName())
+ .append(" sequence similarity matrix\n========\n\n");
+ PrintStream ps = wrapOutputBuffer(sb);
+
+ try
+ {
+ eigenvector = scoreModel.findSimilarities(seqs, similarityParams);
+
+ sb.append(" --- OrigT * Orig ---- \n");
+ eigenvector.print(ps, "%8.2f");
+
+ symm = eigenvector.copy();
+
+ eigenvector.tred();
+
+ sb.append(" ---Tridiag transform matrix ---\n");
+ sb.append(" --- D vector ---\n");
+ eigenvector.printD(ps, "%15.4e");
+ ps.println();
+ sb.append("--- E vector ---\n");
+ eigenvector.printE(ps, "%15.4e");
+ ps.println();
+
+ // Now produce the diagonalization matrix
+ eigenvector.tqli();
+ } catch (Exception q)
+ {
+ q.printStackTrace();
+ sb.append("\n*** Unexpected exception when performing PCA ***\n"
+ + q.getLocalizedMessage());
+ sb.append(
+ "*** Matrices below may not be fully diagonalised. ***\n");
+ }
+
+ sb.append(" --- New diagonalization matrix ---\n");
+ eigenvector.print(ps, "%8.2f");
+ sb.append(" --- Eigenvalues ---\n");
+ eigenvector.printD(ps, "%15.4e");
+ ps.println();
+
+ details = sb.toString();
+ }
+
+ /**
+ * Returns a PrintStream that wraps (sends its output to) the given
+ * StringBuilder
+ *
+ * @param sb
+ * @return
+ */
+ protected PrintStream wrapOutputBuffer(StringBuilder sb)
+ {
+ PrintStream ps = new PrintStream(System.out)
+ {
+ @Override
+ public void print(String x)
+ {
+ sb.append(x);
+ }
+
+ @Override
+ public void println()
+ {
+ sb.append("\n");
+ }
+ };
+ return ps;
+ }
+
+ /**
+ * Answers the N dimensions of the NxN PCA matrix. This is the number of
+ * sequences involved in the pairwise score calculation.
+ *
+ * @return
+ */
+ public int getHeight()
+ {
+ // TODO can any of seqs[] be null?
+ return seqs.getSequences().length;
+ }
+}