Clover icon

Coverage Report

  1. Project Clover database Mon Nov 11 2024 17:27:16 GMT
  2. Package jalview.analysis

File PCA.java

 

Coverage histogram

../../img/srcFileCovDistChart6.png
36% of files have more coverage

Code metrics

6
56
17
1
302
141
21
0.38
3.29
17
1.24

Classes

Class Line # Actions
PCA 35 56 21
0.531645653.2%
 

Contributing tests

This file is covered by 1 test. .

Source view

1    /*
2    * Jalview - A Sequence Alignment Editor and Viewer ($$Version-Rel$$)
3    * Copyright (C) $$Year-Rel$$ The Jalview Authors
4    *
5    * This file is part of Jalview.
6    *
7    * Jalview is free software: you can redistribute it and/or
8    * modify it under the terms of the GNU General Public License
9    * as published by the Free Software Foundation, either version 3
10    * of the License, or (at your option) any later version.
11    *
12    * Jalview is distributed in the hope that it will be useful, but
13    * WITHOUT ANY WARRANTY; without even the implied warranty
14    * of MERCHANTABILITY or FITNESS FOR A PARTICULAR
15    * PURPOSE. See the GNU General Public License for more details.
16    *
17    * You should have received a copy of the GNU General Public License
18    * along with Jalview. If not, see <http://www.gnu.org/licenses/>.
19    * The Jalview Authors are detailed in the 'AUTHORS' file.
20    */
21    package jalview.analysis;
22   
23    import jalview.api.analysis.ScoreModelI;
24    import jalview.api.analysis.SimilarityParamsI;
25    import jalview.bin.Console;
26    import jalview.datamodel.AlignmentView;
27    import jalview.datamodel.Point;
28    import jalview.math.MatrixI;
29   
30    import java.io.PrintStream;
31   
32    /**
33    * Performs Principal Component Analysis on given sequences
34    */
 
35    public class PCA implements Runnable
36    {
37    /*
38    * inputs
39    */
40    final private AlignmentView seqs;
41   
42    final private ScoreModelI scoreModel;
43   
44    final private SimilarityParamsI similarityParams;
45   
46    /*
47    * outputs
48    */
49    private MatrixI pairwiseScores;
50   
51    private MatrixI tridiagonal;
52   
53    private MatrixI eigenMatrix;
54   
55    /**
56    * Constructor given the sequences to compute for, the similarity model to
57    * use, and a set of parameters for sequence comparison
58    *
59    * @param sequences
60    * @param sm
61    * @param options
62    */
 
63  2 toggle public PCA(AlignmentView sequences, ScoreModelI sm,
64    SimilarityParamsI options)
65    {
66  2 this.seqs = sequences;
67  2 this.scoreModel = sm;
68  2 this.similarityParams = options;
69    }
70   
71    /**
72    * Returns Eigenvalue
73    *
74    * @param i
75    * Index of diagonal within matrix
76    *
77    * @return Returns value of diagonal from matrix
78    */
 
79  0 toggle public double getEigenvalue(int i)
80    {
81  0 return eigenMatrix.getD()[i];
82    }
83   
84    /**
85    * DOCUMENT ME!
86    *
87    * @param l
88    * DOCUMENT ME!
89    * @param n
90    * DOCUMENT ME!
91    * @param mm
92    * DOCUMENT ME!
93    * @param factor
94    * DOCUMENT ME!
95    *
96    * @return DOCUMENT ME!
97    */
 
98  1 toggle public Point[] getComponents(int l, int n, int mm, float factor)
99    {
100  1 Point[] out = new Point[getHeight()];
101   
102  16 for (int i = 0; i < getHeight(); i++)
103    {
104  15 float x = (float) component(i, l) * factor;
105  15 float y = (float) component(i, n) * factor;
106  15 float z = (float) component(i, mm) * factor;
107  15 out[i] = new Point(x, y, z);
108    }
109   
110  1 return out;
111    }
112   
113    /**
114    * DOCUMENT ME!
115    *
116    * @param n
117    * DOCUMENT ME!
118    *
119    * @return DOCUMENT ME!
120    */
 
121  0 toggle public double[] component(int n)
122    {
123    // n = index of eigenvector
124  0 double[] out = new double[getHeight()];
125   
126  0 for (int i = 0; i < out.length; i++)
127    {
128  0 out[i] = component(i, n);
129    }
130   
131  0 return out;
132    }
133   
134    /**
135    * DOCUMENT ME!
136    *
137    * @param row
138    * DOCUMENT ME!
139    * @param n
140    * DOCUMENT ME!
141    *
142    * @return DOCUMENT ME!
143    */
 
144  45 toggle double component(int row, int n)
145    {
146  45 double out = 0.0;
147   
148  720 for (int i = 0; i < pairwiseScores.width(); i++)
149    {
150  675 out += (pairwiseScores.getValue(row, i) * eigenMatrix.getValue(i, n));
151    }
152   
153  45 return out / eigenMatrix.getD()[n];
154    }
155   
156    /**
157    * Answers a formatted text report of the PCA calculation results (matrices
158    * and eigenvalues) suitable for display
159    *
160    * @return
161    */
 
162  0 toggle public String getDetails()
163    {
164  0 StringBuilder sb = new StringBuilder(1024);
165  0 sb.append("PCA calculation using ").append(scoreModel.getName())
166    .append(" sequence similarity matrix\n========\n\n");
167  0 PrintStream ps = wrapOutputBuffer(sb);
168   
169    /*
170    * pairwise similarity scores
171    */
172  0 sb.append(" --- OrigT * Orig ---- \n");
173  0 pairwiseScores.print(ps, "%8.2f");
174   
175    /*
176    * tridiagonal matrix, with D and E vectors
177    */
178  0 sb.append(" ---Tridiag transform matrix ---\n");
179  0 sb.append(" --- D vector ---\n");
180  0 tridiagonal.printD(ps, "%15.4e");
181  0 ps.println();
182  0 sb.append("--- E vector ---\n");
183  0 tridiagonal.printE(ps, "%15.4e");
184  0 ps.println();
185   
186    /*
187    * eigenvalues matrix, with D vector
188    */
189  0 sb.append(" --- New diagonalization matrix ---\n");
190  0 eigenMatrix.print(ps, "%8.2f");
191  0 sb.append(" --- Eigenvalues ---\n");
192  0 eigenMatrix.printD(ps, "%15.4e");
193  0 ps.println();
194   
195  0 return sb.toString();
196    }
197   
198    /**
199    * Performs the PCA calculation
200    */
 
201  1 toggle @Override
202    public void run()
203    {
204  1 try
205    {
206    /*
207    * sequence pairwise similarity scores
208    */
209  1 pairwiseScores = scoreModel.findSimilarities(seqs, similarityParams);
210   
211    /*
212    * tridiagonal matrix
213    */
214  1 tridiagonal = pairwiseScores.copy();
215  1 tridiagonal.tred();
216   
217    /*
218    * the diagonalization matrix
219    */
220  1 eigenMatrix = tridiagonal.copy();
221  1 eigenMatrix.tqli();
222    } catch (Exception q)
223    {
224  0 Console.error("Error computing PCA: " + q.getMessage());
225  0 q.printStackTrace();
226    }
227    }
228   
229    /**
230    * Returns a PrintStream that wraps (appends its output to) the given
231    * StringBuilder
232    *
233    * @param sb
234    * @return
235    */
 
236  0 toggle protected PrintStream wrapOutputBuffer(StringBuilder sb)
237    {
238  0 PrintStream ps = new PrintStream(System.out)
239    {
 
240  0 toggle @Override
241    public void print(String x)
242    {
243  0 sb.append(x);
244    }
245   
 
246  0 toggle @Override
247    public void println()
248    {
249  0 sb.append("\n");
250    }
251    };
252  0 return ps;
253    }
254   
255    /**
256    * Answers the N dimensions of the NxN PCA matrix. This is the number of
257    * sequences involved in the pairwise score calculation.
258    *
259    * @return
260    */
 
261  19 toggle public int getHeight()
262    {
263    // TODO can any of seqs[] be null?
264  19 return pairwiseScores.height();// seqs.getSequences().length;
265    }
266   
267    /**
268    * Answers the sequence pairwise similarity scores which were the first step
269    * of the PCA calculation
270    *
271    * @return
272    */
 
273  1 toggle public MatrixI getPairwiseScores()
274    {
275  1 return pairwiseScores;
276    }
277   
 
278  1 toggle public void setPairwiseScores(MatrixI m)
279    {
280  1 pairwiseScores = m;
281    }
282   
 
283  1 toggle public MatrixI getEigenmatrix()
284    {
285  1 return eigenMatrix;
286    }
287   
 
288  1 toggle public void setEigenmatrix(MatrixI m)
289    {
290  1 eigenMatrix = m;
291    }
292   
 
293  1 toggle public MatrixI getTridiagonal()
294    {
295  1 return tridiagonal;
296    }
297   
 
298  1 toggle public void setTridiagonal(MatrixI tridiagonal)
299    {
300  1 this.tridiagonal = tridiagonal;
301    }
302    }