2 * Jalview - A Sequence Alignment Editor and Viewer ($$Version-Rel$$)
3 * Copyright (C) $$Year-Rel$$ The Jalview Authors
5 * This file is part of Jalview.
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.
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.
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.
23 import java.io.PrintStream;
26 * An interface that describes a rectangular matrix of double values and
29 public interface MatrixI
32 * Answers the number of columns
39 * Answers the number of rows
46 * Answers the value at row i, column j
52 double getValue(int i, int j);
55 * Sets the value at row i, colum j
61 void setValue(int i, int j, double d);
64 * Returns the matrix as a double[][] array
71 * Answers a copy of the values in the i'th row
75 double[] getRow(int i);
78 * Answers a copy of the values in the i'th column
82 double[] getColumn(int i);
85 * Answers a new matrix with a copy of the values in this one
92 * Returns a new matrix which is the transpose of this one
99 * Returns a new matrix which is the result of premultiplying this matrix by
100 * the supplied argument. If this of size AxB (A rows and B columns), and the
101 * argument is CxA (C rows and A columns), the result is of size CxB.
106 * @throws IllegalArgumentException
107 * if the number of columns in the pre-multiplier is not equal to
108 * the number of rows in the multiplicand (this)
110 MatrixI preMultiply(MatrixI m);
113 * Returns a new matrix which is the result of postmultiplying this matrix by
114 * the supplied argument. If this of size AxB (A rows and B columns), and the
115 * argument is BxC (B rows and C columns), the result is of size AxC.
117 * This method simply returns the result of in.preMultiply(this)
122 * @throws IllegalArgumentException
123 * if the number of rows in the post-multiplier is not equal to the
124 * number of columns in the multiplicand (this)
125 * @see #preMultiply(Matrix)
127 MatrixI postMultiply(MatrixI m);
133 void setD(double[] v);
135 void setE(double[] v);
137 void print(PrintStream ps, String format);
139 void printD(PrintStream ps, String format);
141 void printE(PrintStream ps, String format);
143 void tqli() throws Exception;
148 * Reverses the range of the matrix values, so that the smallest values become
149 * the largest, and the largest become the smallest. This operation supports
150 * using a distance measure as a similarity measure, or vice versa.
152 * If parameter <code>maxToZero</code> is true, then the maximum value becomes
153 * zero, i.e. all values are subtracted from the maximum. This is consistent
154 * with converting an identity similarity score to a distance score - the most
155 * similar (identity) corresponds to zero distance. However note that the
156 * operation is not reversible (unless the original minimum value is zero).
157 * For example a range of 10-40 would become 30-0, which would reverse a
158 * second time to 0-30. Also note that a general similarity measure (such as
159 * BLOSUM) may give different 'identity' scores for different sequences, so
160 * they cannot all convert to zero distance.
162 * If parameter <code>maxToZero</code> is false, then the values are reflected
163 * about the average of {min, max} (effectively swapping min and max). This
164 * operation <em>is</em> reversible.
168 void reverseRange(boolean maxToZero);
171 * Multiply all entries by the given value
175 void multiply(double d);
178 * Add d to all entries of this matrix
180 * @param d ~ value to add
185 * Answers true if the two matrices have the same dimensions, and
186 * corresponding values all differ by no more than delta (which should be a
187 * positive value), else false
193 boolean equals(MatrixI m2, double delta);
196 * Returns a copy in which every value in the matrix is its absolute
201 * Returns the mean of each row
206 * Returns the mean of each column
211 * Returns a flattened matrix containing the sum of each column
218 * returns the mean value of the complete matrix
223 * fills up a diagonal matrix with its transposed copy
224 * !other side should be filled with either 0 or Double.NaN
229 * counts the number of Double.NaN in the matrix
236 * performs an element-wise addition of this matrix by another matrix
237 * !matrices have to be the same size
238 * @param m ~ other matrix
241 * @throws IllegalArgumentException
242 * if this and m do not have the same dimensions
244 MatrixI add(MatrixI m);
247 * performs an element-wise subtraction of this matrix by another matrix
248 * !matrices have to be the same size
249 * @param m ~ other matrix
252 * @throws IllegalArgumentException
253 * if this and m do not have the same dimensions
255 MatrixI subtract(MatrixI m);
258 * performs an element-wise multiplication of this matrix by another matrix ~ this * m
259 * !matrices have to be the same size
260 * @param m ~ other matrix
263 * @throws IllegalArgumentException
264 * if this and m do not have the same dimensions
266 MatrixI elementwiseMultiply(MatrixI m);
269 * performs an element-wise division of this matrix by another matrix ~ this / m
270 * !matrices have to be the same size
271 * @param m ~ other matrix
274 * @throws IllegalArgumentException
275 * if this and m do not have the same dimensions
277 MatrixI elementwiseDivide(MatrixI m);
280 * calculates the root-mean-square for two matrices
281 * @param m ~ other matrix
285 double rmsd(MatrixI m);
288 * calculates the Frobenius norm of this matrix
295 * returns the sum of all values in this matrix
302 * returns the sum-product of this matrix with vector v
306 * @throws IllegalArgumentException
307 * if this.cols and v do not have the same length
309 double[] sumProduct(double[] v);
312 * mirrors the columns of this matrix