1、特征值分解基礎知識

矩陣乘法Y=AB的數學意義在于變換，以其中一個向量A為中心，則B的作用主要是使A發生伸縮或旋轉變換。一個矩陣其實就是一個線性變換，因為一個矩陣乘以一個向量后得到的向量，其實就相當于將這個向量進行了線性變換。

如果說一個向量v是方陣A的特征向量，將一定可以表示成下面的形式：

??? 這時候λ就被稱為特征向量v對應的特征值，一個矩陣的一組特征向量是一組正交向量。特征值分解是將一個矩陣分解成下面的形式：

其中Q是這個矩陣A的特征向量組成的矩陣，Σ是一個對角陣，每一個對角線上的元素就是一個特征值。一個變換方陣的所有特征向量組成了這個變換矩陣的一組基。所謂基，可以理解為坐標系的軸。平常用到直角坐標系，在線性代數中可以把這個坐標系扭曲、拉伸、旋轉，稱為基變換。可以按需求去設定基，但是基的軸之間必須是線性無關的，也就是保證坐標系的不同軸不要指向同一個方向或可以被別的軸組合而成。從線性空間的角度看，在一個定義了內積的線性空間里，對一個N階對稱方陣進行特征分解，就是產生了該空間的N個標準正交基，然后把矩陣投影到這N個基上。N個特征向量就是N個標準正交基，而特征值的模則代表矩陣在每個基上的投影長度。特征值越大，說明矩陣在對應的特征向量上的方差越大，功率越大，信息量越多。不過，特征值分解也有很多的局限，比如說變換的矩陣必須是方陣。

在機器學習特征提取中，意思就是最大特征值對應的特征向量方向上包含最多的信息量，如果某幾個特征值很小，說明這幾個方向信息量很小，可以用來降維，也就是刪除小特征值對應方向的數據，只保留大特征值方向對應的數據，這樣做以后數據量減小，但有用信息量變化不大，PCA降維就是基于這種思路。特征值分解可以得到特征值與特征向量，特征值表示的是這個特征到底有多重要，而特征向量表示這個特征是什么，可以將每一個特征向量理解為一個線性的子空間。

2、Java實現

http://math.nist.gov/javanumerics/jama/Java矩陣計算包，下載Jama-1.0.3.jar引入工程。

下載Jama-1.0.3.zip研究源碼。

1）? 特征值分解測試類

?

package sk.ml;import Jama.EigenvalueDecomposition; import Jama.Matrix;public class QRTest {//矩陣特征分解public static void main(String argv[]){double[] columnwise = {1.,2.,3.,4.,5.,6.,7.,8.,9.,10.,11.,12.};Matrix A = new Matrix(columnwise,4);//構造矩陣A.print(A.getColumnDimension(), A.getRowDimension());EigenvalueDecomposition Eig = A.eig();Matrix D = Eig.getD();Matrix V = Eig.getV();D.print(D.getColumnDimension(), D.getRowDimension());//打印特征值V.print(V.getColumnDimension(), V.getRowDimension());//打印特征向量} }

?

2）? 源碼參考Matrix

?

package Jama;import java.text.NumberFormat; import java.text.DecimalFormat; import java.text.DecimalFormatSymbols; import java.util.Locale; import java.text.FieldPosition; import java.io.PrintWriter; import java.io.BufferedReader; import java.io.StreamTokenizer; import Jama.util.*;/**Jama = Java Matrix class. <P>The Java Matrix Class provides the fundamental operations of numericallinear algebra. Various constructors create Matrices from two dimensionalarrays of double precision floating point numbers. Various "gets" and"sets" provide access to submatrices and matrix elements. Several methods implement basic matrix arithmetic, including matrix addition andmultiplication, matrix norms, and element-by-element array operations.Methods for reading and printing matrices are also included. All theoperations in this version of the Matrix Class involve real matrices.Complex matrices may be handled in a future version. <P>Five fundamental matrix decompositions, which consist of pairs or triplesof matrices, permutation vectors, and the like, produce results in fivedecomposition classes. These decompositions are accessed by the Matrixclass to compute solutions of simultaneous linear equations, determinants,inverses and other matrix functions. The five decompositions are: <P><UL><LI>Cholesky Decomposition of symmetric, positive definite matrices.<LI>LU Decomposition of rectangular matrices.<LI>QR Decomposition of rectangular matrices.<LI>Singular Value Decomposition of rectangular matrices.<LI>Eigenvalue Decomposition of both symmetric and nonsymmetric square matrices. </UL> <DL> <DT><B>Example of use:</B></DT> <P> <DD>Solve a linear system A x = b and compute the residual norm, ||b - A x||. <P><PRE>double[][] vals = {{1.,2.,3},{4.,5.,6.},{7.,8.,10.}};Matrix A = new Matrix(vals);Matrix b = Matrix.random(3,1);Matrix x = A.solve(b);Matrix r = A.times(x).minus(b);double rnorm = r.normInf(); </PRE></DD> </DL>@author The MathWorks, Inc. and the National Institute of Standards and Technology. @version 5 August 1998 */public class Matrix implements Cloneable, java.io.Serializable {/* ------------------------Class variables* ------------------------ *//** Array for internal storage of elements.@serial internal array storage.*/private double[][] A;/** Row and column dimensions.@serial row dimension.@serial column dimension.*/private int m, n;/* ------------------------Constructors* ------------------------ *//** Construct an m-by-n matrix of zeros. @param m Number of rows.@param n Number of colums.*/public Matrix (int m, int n) {this.m = m;this.n = n;A = new double[m][n];}/** Construct an m-by-n constant matrix.@param m Number of rows.@param n Number of colums.@param s Fill the matrix with this scalar value.*/public Matrix (int m, int n, double s) {this.m = m;this.n = n;A = new double[m][n];for (int i = 0; i < m; i++) {for (int j = 0; j < n; j++) {A[i][j] = s;}}}/** Construct a matrix from a 2-D array.@param A Two-dimensional array of doubles.@exception IllegalArgumentException All rows must have the same length@see #constructWithCopy*/public Matrix (double[][] A) {m = A.length;n = A[0].length;for (int i = 0; i < m; i++) {if (A[i].length != n) {throw new IllegalArgumentException("All rows must have the same length.");}}this.A = A;}/** Construct a matrix quickly without checking arguments.@param A Two-dimensional array of doubles.@param m Number of rows.@param n Number of colums.*/public Matrix (double[][] A, int m, int n) {this.A = A;this.m = m;this.n = n;}/** Construct a matrix from a one-dimensional packed array@param vals One-dimensional array of doubles, packed by columns (ala Fortran).@param m Number of rows.@exception IllegalArgumentException Array length must be a multiple of m.*/public Matrix (double vals[], int m) {this.m = m;n = (m != 0 ? vals.length/m : 0);if (m*n != vals.length) {throw new IllegalArgumentException("Array length must be a multiple of m.");}A = new double[m][n];for (int i = 0; i < m; i++) {for (int j = 0; j < n; j++) {A[i][j] = vals[i+j*m];}}}/* ------------------------Public Methods* ------------------------ *//** Construct a matrix from a copy of a 2-D array.@param A Two-dimensional array of doubles.@exception IllegalArgumentException All rows must have the same length*/public static Matrix constructWithCopy(double[][] A) {int m = A.length;int n = A[0].length;Matrix X = new Matrix(m,n);double[][] C = X.getArray();for (int i = 0; i < m; i++) {if (A[i].length != n) {throw new IllegalArgumentException("All rows must have the same length.");}for (int j = 0; j < n; j++) {C[i][j] = A[i][j];}}return X;}/** Make a deep copy of a matrix*/public Matrix copy () {Matrix X = new Matrix(m,n);double[][] C = X.getArray();for (int i = 0; i < m; i++) {for (int j = 0; j < n; j++) {C[i][j] = A[i][j];}}return X;}/** Clone the Matrix object.*/public Object clone () {return this.copy();}/** Access the internal two-dimensional array.@return Pointer to the two-dimensional array of matrix elements.*/public double[][] getArray () {return A;}/** Copy the internal two-dimensional array.@return Two-dimensional array copy of matrix elements.*/public double[][] getArrayCopy () {double[][] C = new double[m][n];for (int i = 0; i < m; i++) {for (int j = 0; j < n; j++) {C[i][j] = A[i][j];}}return C;}/** Make a one-dimensional column packed copy of the internal array.@return Matrix elements packed in a one-dimensional array by columns.*/public double[] getColumnPackedCopy () {double[] vals = new double[m*n];for (int i = 0; i < m; i++) {for (int j = 0; j < n; j++) {vals[i+j*m] = A[i][j];}}return vals;}/** Make a one-dimensional row packed copy of the internal array.@return Matrix elements packed in a one-dimensional array by rows.*/public double[] getRowPackedCopy () {double[] vals = new double[m*n];for (int i = 0; i < m; i++) {for (int j = 0; j < n; j++) {vals[i*n+j] = A[i][j];}}return vals;}/** Get row dimension.@return m, the number of rows.*/public int getRowDimension () {return m;}/** Get column dimension.@return n, the number of columns.*/public int getColumnDimension () {return n;}/** Get a single element.@param i Row index.@param j Column index.@return A(i,j)@exception ArrayIndexOutOfBoundsException*/public double get (int i, int j) {return A[i][j];}/** Get a submatrix.@param i0 Initial row index@param i1 Final row index@param j0 Initial column index@param j1 Final column index@return A(i0:i1,j0:j1)@exception ArrayIndexOutOfBoundsException Submatrix indices*/public Matrix getMatrix (int i0, int i1, int j0, int j1) {Matrix X = new Matrix(i1-i0+1,j1-j0+1);double[][] B = X.getArray();try {for (int i = i0; i <= i1; i++) {for (int j = j0; j <= j1; j++) {B[i-i0][j-j0] = A[i][j];}}} catch(ArrayIndexOutOfBoundsException e) {throw new ArrayIndexOutOfBoundsException("Submatrix indices");}return X;}/** Get a submatrix.@param r Array of row indices.@param c Array of column indices.@return A(r(:),c(:))@exception ArrayIndexOutOfBoundsException Submatrix indices*/public Matrix getMatrix (int[] r, int[] c) {Matrix X = new Matrix(r.length,c.length);double[][] B = X.getArray();try {for (int i = 0; i < r.length; i++) {for (int j = 0; j < c.length; j++) {B[i][j] = A[r[i]][c[j]];}}} catch(ArrayIndexOutOfBoundsException e) {throw new ArrayIndexOutOfBoundsException("Submatrix indices");}return X;}/** Get a submatrix.@param i0 Initial row index@param i1 Final row index@param c Array of column indices.@return A(i0:i1,c(:))@exception ArrayIndexOutOfBoundsException Submatrix indices*/public Matrix getMatrix (int i0, int i1, int[] c) {Matrix X = new Matrix(i1-i0+1,c.length);double[][] B = X.getArray();try {for (int i = i0; i <= i1; i++) {for (int j = 0; j < c.length; j++) {B[i-i0][j] = A[i][c[j]];}}} catch(ArrayIndexOutOfBoundsException e) {throw new ArrayIndexOutOfBoundsException("Submatrix indices");}return X;}/** Get a submatrix.@param r Array of row indices.@param j0 Initial column index@param j1 Final column index@return A(r(:),j0:j1)@exception ArrayIndexOutOfBoundsException Submatrix indices*/public Matrix getMatrix (int[] r, int j0, int j1) {Matrix X = new Matrix(r.length,j1-j0+1);double[][] B = X.getArray();try {for (int i = 0; i < r.length; i++) {for (int j = j0; j <= j1; j++) {B[i][j-j0] = A[r[i]][j];}}} catch(ArrayIndexOutOfBoundsException e) {throw new ArrayIndexOutOfBoundsException("Submatrix indices");}return X;}/** Set a single element.@param i Row index.@param j Column index.@param s A(i,j).@exception ArrayIndexOutOfBoundsException*/public void set (int i, int j, double s) {A[i][j] = s;}/** Set a submatrix.@param i0 Initial row index@param i1 Final row index@param j0 Initial column index@param j1 Final column index@param X A(i0:i1,j0:j1)@exception ArrayIndexOutOfBoundsException Submatrix indices*/public void setMatrix (int i0, int i1, int j0, int j1, Matrix X) {try {for (int i = i0; i <= i1; i++) {for (int j = j0; j <= j1; j++) {A[i][j] = X.get(i-i0,j-j0);}}} catch(ArrayIndexOutOfBoundsException e) {throw new ArrayIndexOutOfBoundsException("Submatrix indices");}}/** Set a submatrix.@param r Array of row indices.@param c Array of column indices.@param X A(r(:),c(:))@exception ArrayIndexOutOfBoundsException Submatrix indices*/public void setMatrix (int[] r, int[] c, Matrix X) {try {for (int i = 0; i < r.length; i++) {for (int j = 0; j < c.length; j++) {A[r[i]][c[j]] = X.get(i,j);}}} catch(ArrayIndexOutOfBoundsException e) {throw new ArrayIndexOutOfBoundsException("Submatrix indices");}}/** Set a submatrix.@param r Array of row indices.@param j0 Initial column index@param j1 Final column index@param X A(r(:),j0:j1)@exception ArrayIndexOutOfBoundsException Submatrix indices*/public void setMatrix (int[] r, int j0, int j1, Matrix X) {try {for (int i = 0; i < r.length; i++) {for (int j = j0; j <= j1; j++) {A[r[i]][j] = X.get(i,j-j0);}}} catch(ArrayIndexOutOfBoundsException e) {throw new ArrayIndexOutOfBoundsException("Submatrix indices");}}/** Set a submatrix.@param i0 Initial row index@param i1 Final row index@param c Array of column indices.@param X A(i0:i1,c(:))@exception ArrayIndexOutOfBoundsException Submatrix indices*/public void setMatrix (int i0, int i1, int[] c, Matrix X) {try {for (int i = i0; i <= i1; i++) {for (int j = 0; j < c.length; j++) {A[i][c[j]] = X.get(i-i0,j);}}} catch(ArrayIndexOutOfBoundsException e) {throw new ArrayIndexOutOfBoundsException("Submatrix indices");}}/** Matrix transpose.@return A'*/public Matrix transpose () {Matrix X = new Matrix(n,m);double[][] C = X.getArray();for (int i = 0; i < m; i++) {for (int j = 0; j < n; j++) {C[j][i] = A[i][j];}}return X;}/** One norm@return maximum column sum.*/public double norm1 () {double f = 0;for (int j = 0; j < n; j++) {double s = 0;for (int i = 0; i < m; i++) {s += Math.abs(A[i][j]);}f = Math.max(f,s);}return f;}/** Two norm@return maximum singular value.*/public double norm2 () {return (new SingularValueDecomposition(this).norm2());}/** Infinity norm@return maximum row sum.*/public double normInf () {double f = 0;for (int i = 0; i < m; i++) {double s = 0;for (int j = 0; j < n; j++) {s += Math.abs(A[i][j]);}f = Math.max(f,s);}return f;}/** Frobenius norm@return sqrt of sum of squares of all elements.*/public double normF () {double f = 0;for (int i = 0; i < m; i++) {for (int j = 0; j < n; j++) {f = Maths.hypot(f,A[i][j]);}}return f;}/** Unary minus@return -A*/public Matrix uminus () {Matrix X = new Matrix(m,n);double[][] C = X.getArray();for (int i = 0; i < m; i++) {for (int j = 0; j < n; j++) {C[i][j] = -A[i][j];}}return X;}/** C = A + B@param B another matrix@return A + B*/public Matrix plus (Matrix B) {checkMatrixDimensions(B);Matrix X = new Matrix(m,n);double[][] C = X.getArray();for (int i = 0; i < m; i++) {for (int j = 0; j < n; j++) {C[i][j] = A[i][j] + B.A[i][j];}}return X;}/** A = A + B@param B another matrix@return A + B*/public Matrix plusEquals (Matrix B) {checkMatrixDimensions(B);for (int i = 0; i < m; i++) {for (int j = 0; j < n; j++) {A[i][j] = A[i][j] + B.A[i][j];}}return this;}/** C = A - B@param B another matrix@return A - B*/public Matrix minus (Matrix B) {checkMatrixDimensions(B);Matrix X = new Matrix(m,n);double[][] C = X.getArray();for (int i = 0; i < m; i++) {for (int j = 0; j < n; j++) {C[i][j] = A[i][j] - B.A[i][j];}}return X;}/** A = A - B@param B another matrix@return A - B*/public Matrix minusEquals (Matrix B) {checkMatrixDimensions(B);for (int i = 0; i < m; i++) {for (int j = 0; j < n; j++) {A[i][j] = A[i][j] - B.A[i][j];}}return this;}/** Element-by-element multiplication, C = A.*B@param B another matrix@return A.*B*/public Matrix arrayTimes (Matrix B) {checkMatrixDimensions(B);Matrix X = new Matrix(m,n);double[][] C = X.getArray();for (int i = 0; i < m; i++) {for (int j = 0; j < n; j++) {C[i][j] = A[i][j] * B.A[i][j];}}return X;}/** Element-by-element multiplication in place, A = A.*B@param B another matrix@return A.*B*/public Matrix arrayTimesEquals (Matrix B) {checkMatrixDimensions(B);for (int i = 0; i < m; i++) {for (int j = 0; j < n; j++) {A[i][j] = A[i][j] * B.A[i][j];}}return this;}/** Element-by-element right division, C = A./B@param B another matrix@return A./B*/public Matrix arrayRightDivide (Matrix B) {checkMatrixDimensions(B);Matrix X = new Matrix(m,n);double[][] C = X.getArray();for (int i = 0; i < m; i++) {for (int j = 0; j < n; j++) {C[i][j] = A[i][j] / B.A[i][j];}}return X;}/** Element-by-element right division in place, A = A./B@param B another matrix@return A./B*/public Matrix arrayRightDivideEquals (Matrix B) {checkMatrixDimensions(B);for (int i = 0; i < m; i++) {for (int j = 0; j < n; j++) {A[i][j] = A[i][j] / B.A[i][j];}}return this;}/** Element-by-element left division, C = A.\B@param B another matrix@return A.\B*/public Matrix arrayLeftDivide (Matrix B) {checkMatrixDimensions(B);Matrix X = new Matrix(m,n);double[][] C = X.getArray();for (int i = 0; i < m; i++) {for (int j = 0; j < n; j++) {C[i][j] = B.A[i][j] / A[i][j];}}return X;}/** Element-by-element left division in place, A = A.\B@param B another matrix@return A.\B*/public Matrix arrayLeftDivideEquals (Matrix B) {checkMatrixDimensions(B);for (int i = 0; i < m; i++) {for (int j = 0; j < n; j++) {A[i][j] = B.A[i][j] / A[i][j];}}return this;}/** Multiply a matrix by a scalar, C = s*A@param s scalar@return s*A*/public Matrix times (double s) {Matrix X = new Matrix(m,n);double[][] C = X.getArray();for (int i = 0; i < m; i++) {for (int j = 0; j < n; j++) {C[i][j] = s*A[i][j];}}return X;}/** Multiply a matrix by a scalar in place, A = s*A@param s scalar@return replace A by s*A*/public Matrix timesEquals (double s) {for (int i = 0; i < m; i++) {for (int j = 0; j < n; j++) {A[i][j] = s*A[i][j];}}return this;}/** Linear algebraic matrix multiplication, A * B@param B another matrix@return Matrix product, A * B@exception IllegalArgumentException Matrix inner dimensions must agree.*/public Matrix times (Matrix B) {if (B.m != n) {throw new IllegalArgumentException("Matrix inner dimensions must agree.");}Matrix X = new Matrix(m,B.n);double[][] C = X.getArray();double[] Bcolj = new double[n];for (int j = 0; j < B.n; j++) {for (int k = 0; k < n; k++) {Bcolj[k] = B.A[k][j];}for (int i = 0; i < m; i++) {double[] Arowi = A[i];double s = 0;for (int k = 0; k < n; k++) {s += Arowi[k]*Bcolj[k];}C[i][j] = s;}}return X;}/** LU Decomposition@return LUDecomposition@see LUDecomposition*/public LUDecomposition lu () {return new LUDecomposition(this);}/** QR Decomposition@return QRDecomposition@see QRDecomposition*/public QRDecomposition qr () {return new QRDecomposition(this);}/** Cholesky Decomposition@return CholeskyDecomposition@see CholeskyDecomposition*/public CholeskyDecomposition chol () {return new CholeskyDecomposition(this);}/** Singular Value Decomposition@return SingularValueDecomposition@see SingularValueDecomposition*/public SingularValueDecomposition svd () {return new SingularValueDecomposition(this);}/** Eigenvalue Decomposition@return EigenvalueDecomposition@see EigenvalueDecomposition*/public EigenvalueDecomposition eig () {return new EigenvalueDecomposition(this);}/** Solve A*X = B@param B right hand side@return solution if A is square, least squares solution otherwise*/public Matrix solve (Matrix B) {return (m == n ? (new LUDecomposition(this)).solve(B) :(new QRDecomposition(this)).solve(B));}/** Solve X*A = B, which is also A'*X' = B'@param B right hand side@return solution if A is square, least squares solution otherwise.*/public Matrix solveTranspose (Matrix B) {return transpose().solve(B.transpose());}/** Matrix inverse or pseudoinverse@return inverse(A) if A is square, pseudoinverse otherwise.*/public Matrix inverse () {return solve(identity(m,m));}/** Matrix determinant@return determinant*/public double det () {return new LUDecomposition(this).det();}/** Matrix rank@return effective numerical rank, obtained from SVD.*/public int rank () {return new SingularValueDecomposition(this).rank();}/** Matrix condition (2 norm)@return ratio of largest to smallest singular value.*/public double cond () {return new SingularValueDecomposition(this).cond();}/** Matrix trace.@return sum of the diagonal elements.*/public double trace () {double t = 0;for (int i = 0; i < Math.min(m,n); i++) {t += A[i][i];}return t;}/** Generate matrix with random elements@param m Number of rows.@param n Number of colums.@return An m-by-n matrix with uniformly distributed random elements.*/public static Matrix random (int m, int n) {Matrix A = new Matrix(m,n);double[][] X = A.getArray();for (int i = 0; i < m; i++) {for (int j = 0; j < n; j++) {X[i][j] = Math.random();}}return A;}/** Generate identity matrix@param m Number of rows.@param n Number of colums.@return An m-by-n matrix with ones on the diagonal and zeros elsewhere.*/public static Matrix identity (int m, int n) {Matrix A = new Matrix(m,n);double[][] X = A.getArray();for (int i = 0; i < m; i++) {for (int j = 0; j < n; j++) {X[i][j] = (i == j ? 1.0 : 0.0);}}return A;}/** Print the matrix to stdout. Line the elements up in columns* with a Fortran-like 'Fw.d' style format.@param w Column width.@param d Number of digits after the decimal.*/public void print (int w, int d) {print(new PrintWriter(System.out,true),w,d); }/** Print the matrix to the output stream. Line the elements up in* columns with a Fortran-like 'Fw.d' style format.@param output Output stream.@param w Column width.@param d Number of digits after the decimal.*/public void print (PrintWriter output, int w, int d) {DecimalFormat format = new DecimalFormat();format.setDecimalFormatSymbols(new DecimalFormatSymbols(Locale.US));format.setMinimumIntegerDigits(1);format.setMaximumFractionDigits(d);format.setMinimumFractionDigits(d);format.setGroupingUsed(false);print(output,format,w+2);}/** Print the matrix to stdout. Line the elements up in columns.* Use the format object, and right justify within columns of width* characters.* Note that is the matrix is to be read back in, you probably will want* to use a NumberFormat that is set to US Locale.@param format A Formatting object for individual elements.@param width Field width for each column.@see java.text.DecimalFormat#setDecimalFormatSymbols*/public void print (NumberFormat format, int width) {print(new PrintWriter(System.out,true),format,width); }// DecimalFormat is a little disappointing coming from Fortran or C's printf.// Since it doesn't pad on the left, the elements will come out different// widths. Consequently, we'll pass the desired column width in as an// argument and do the extra padding ourselves./** Print the matrix to the output stream. Line the elements up in columns.* Use the format object, and right justify within columns of width* characters.* Note that is the matrix is to be read back in, you probably will want* to use a NumberFormat that is set to US Locale.@param output the output stream.@param format A formatting object to format the matrix elements @param width Column width.@see java.text.DecimalFormat#setDecimalFormatSymbols*/public void print (PrintWriter output, NumberFormat format, int width) {output.println(); // start on new line.for (int i = 0; i < m; i++) {for (int j = 0; j < n; j++) {String s = format.format(A[i][j]); // format the numberint padding = Math.max(1,width-s.length()); // At _least_ 1 spacefor (int k = 0; k < padding; k++)output.print(' ');output.print(s);}output.println();}output.println(); // end with blank line.}/** Read a matrix from a stream. The format is the same the print method,* so printed matrices can be read back in (provided they were printed using* US Locale). Elements are separated by* whitespace, all the elements for each row appear on a single line,* the last row is followed by a blank line.@param input the input stream.*/public static Matrix read (BufferedReader input) throws java.io.IOException {StreamTokenizer tokenizer= new StreamTokenizer(input);// Although StreamTokenizer will parse numbers, it doesn't recognize// scientific notation (E or D); however, Double.valueOf does.// The strategy here is to disable StreamTokenizer's number parsing.// We'll only get whitespace delimited words, EOL's and EOF's.// These words should all be numbers, for Double.valueOf to parse.tokenizer.resetSyntax();tokenizer.wordChars(0,255);tokenizer.whitespaceChars(0, ' ');tokenizer.eolIsSignificant(true);java.util.Vector<Double> vD = new java.util.Vector<Double>();// Ignore initial empty lineswhile (tokenizer.nextToken() == StreamTokenizer.TT_EOL);if (tokenizer.ttype == StreamTokenizer.TT_EOF)throw new java.io.IOException("Unexpected EOF on matrix read.");do {vD.addElement(Double.valueOf(tokenizer.sval)); // Read & store 1st row.} while (tokenizer.nextToken() == StreamTokenizer.TT_WORD);int n = vD.size(); // Now we've got the number of columns!double row[] = new double[n];for (int j=0; j<n; j++) // extract the elements of the 1st row.row[j]=vD.elementAt(j).doubleValue();java.util.Vector<double[]> v = new java.util.Vector<double[]>();v.addElement(row); // Start storing rows instead of columns.while (tokenizer.nextToken() == StreamTokenizer.TT_WORD) {// While non-empty linesv.addElement(row = new double[n]);int j = 0;do {if (j >= n) throw new java.io.IOException("Row " + v.size() + " is too long.");row[j++] = Double.valueOf(tokenizer.sval).doubleValue();} while (tokenizer.nextToken() == StreamTokenizer.TT_WORD);if (j < n) throw new java.io.IOException("Row " + v.size() + " is too short.");}int m = v.size(); // Now we've got the number of rows.double[][] A = new double[m][];v.copyInto(A); // copy the rows out of the vectorreturn new Matrix(A);}/* ------------------------Private Methods* ------------------------ *//** Check if size(A) == size(B) **/private void checkMatrixDimensions (Matrix B) {if (B.m != m || B.n != n) {throw new IllegalArgumentException("Matrix dimensions must agree.");}}private static final long serialVersionUID = 1; }

?

3）源碼參考EigenvalueDecomposition

可重點研讀如何實現特征值分解。

package Jama; import Jama.util.*;/** Eigenvalues and eigenvectors of a real matrix. <P>If A is symmetric, then A = V*D*V' where the eigenvalue matrix D isdiagonal and the eigenvector matrix V is orthogonal.I.e. A = V.times(D.times(V.transpose())) and V.times(V.transpose()) equals the identity matrix. <P>If A is not symmetric, then the eigenvalue matrix D is block diagonalwith the real eigenvalues in 1-by-1 blocks and any complex eigenvalues,lambda + i*mu, in 2-by-2 blocks, [lambda, mu; -mu, lambda]. Thecolumns of V represent the eigenvectors in the sense that A*V = V*D,i.e. A.times(V) equals V.times(D). The matrix V may be badlyconditioned, or even singular, so the validity of the equationA = V*D*inverse(V) depends upon V.cond(). **/public class EigenvalueDecomposition implements java.io.Serializable {/* ------------------------Class variables* ------------------------ *//** Row and column dimension (square matrix).@serial matrix dimension.*/private int n;/** Symmetry flag.@serial internal symmetry flag.*/private boolean issymmetric;/** Arrays for internal storage of eigenvalues.@serial internal storage of eigenvalues.*/private double[] d, e;/** Array for internal storage of eigenvectors.@serial internal storage of eigenvectors.*/private double[][] V;/** Array for internal storage of nonsymmetric Hessenberg form.@serial internal storage of nonsymmetric Hessenberg form.*/private double[][] H;/** Working storage for nonsymmetric algorithm.@serial working storage for nonsymmetric algorithm.*/private double[] ort;/* ------------------------Private Methods* ------------------------ */// Symmetric Householder reduction to tridiagonal form.private void tred2 () {// This is derived from the Algol procedures tred2 by// Bowdler, Martin, Reinsch, and Wilkinson, Handbook for// Auto. Comp., Vol.ii-Linear Algebra, and the corresponding// Fortran subroutine in EISPACK.for (int j = 0; j < n; j++) {d[j] = V[n-1][j];}// Householder reduction to tridiagonal form.for (int i = n-1; i > 0; i--) {// Scale to avoid under/overflow.double scale = 0.0;double h = 0.0;for (int k = 0; k < i; k++) {scale = scale + Math.abs(d[k]);}if (scale == 0.0) {e[i] = d[i-1];for (int j = 0; j < i; j++) {d[j] = V[i-1][j];V[i][j] = 0.0;V[j][i] = 0.0;}} else {// Generate Householder vector.for (int k = 0; k < i; k++) {d[k] /= scale;h += d[k] * d[k];}double f = d[i-1];double g = Math.sqrt(h);if (f > 0) {g = -g;}e[i] = scale * g;h = h - f * g;d[i-1] = f - g;for (int j = 0; j < i; j++) {e[j] = 0.0;}// Apply similarity transformation to remaining columns.for (int j = 0; j < i; j++) {f = d[j];V[j][i] = f;g = e[j] + V[j][j] * f;for (int k = j+1; k <= i-1; k++) {g += V[k][j] * d[k];e[k] += V[k][j] * f;}e[j] = g;}f = 0.0;for (int j = 0; j < i; j++) {e[j] /= h;f += e[j] * d[j];}double hh = f / (h + h);for (int j = 0; j < i; j++) {e[j] -= hh * d[j];}for (int j = 0; j < i; j++) {f = d[j];g = e[j];for (int k = j; k <= i-1; k++) {V[k][j] -= (f * e[k] + g * d[k]);}d[j] = V[i-1][j];V[i][j] = 0.0;}}d[i] = h;}// Accumulate transformations.for (int i = 0; i < n-1; i++) {V[n-1][i] = V[i][i];V[i][i] = 1.0;double h = d[i+1];if (h != 0.0) {for (int k = 0; k <= i; k++) {d[k] = V[k][i+1] / h;}for (int j = 0; j <= i; j++) {double g = 0.0;for (int k = 0; k <= i; k++) {g += V[k][i+1] * V[k][j];}for (int k = 0; k <= i; k++) {V[k][j] -= g * d[k];}}}for (int k = 0; k <= i; k++) {V[k][i+1] = 0.0;}}for (int j = 0; j < n; j++) {d[j] = V[n-1][j];V[n-1][j] = 0.0;}V[n-1][n-1] = 1.0;e[0] = 0.0;} // Symmetric tridiagonal QL algorithm.private void tql2 () {// This is derived from the Algol procedures tql2, by// Bowdler, Martin, Reinsch, and Wilkinson, Handbook for// Auto. Comp., Vol.ii-Linear Algebra, and the corresponding// Fortran subroutine in EISPACK.for (int i = 1; i < n; i++) {e[i-1] = e[i];}e[n-1] = 0.0;double f = 0.0;double tst1 = 0.0;double eps = Math.pow(2.0,-52.0);for (int l = 0; l < n; l++) {// Find small subdiagonal elementtst1 = Math.max(tst1,Math.abs(d[l]) + Math.abs(e[l]));int m = l;while (m < n) {if (Math.abs(e[m]) <= eps*tst1) {break;}m++;}// If m == l, d[l] is an eigenvalue,// otherwise, iterate.if (m > l) {int iter = 0;do {iter = iter + 1; // (Could check iteration count here.)// Compute implicit shiftdouble g = d[l];double p = (d[l+1] - g) / (2.0 * e[l]);double r = Maths.hypot(p,1.0);if (p < 0) {r = -r;}d[l] = e[l] / (p + r);d[l+1] = e[l] * (p + r);double dl1 = d[l+1];double h = g - d[l];for (int i = l+2; i < n; i++) {d[i] -= h;}f = f + h;// Implicit QL transformation.p = d[m];double c = 1.0;double c2 = c;double c3 = c;double el1 = e[l+1];double s = 0.0;double s2 = 0.0;for (int i = m-1; i >= l; i--) {c3 = c2;c2 = c;s2 = s;g = c * e[i];h = c * p;r = Maths.hypot(p,e[i]);e[i+1] = s * r;s = e[i] / r;c = p / r;p = c * d[i] - s * g;d[i+1] = h + s * (c * g + s * d[i]);// Accumulate transformation.for (int k = 0; k < n; k++) {h = V[k][i+1];V[k][i+1] = s * V[k][i] + c * h;V[k][i] = c * V[k][i] - s * h;}}p = -s * s2 * c3 * el1 * e[l] / dl1;e[l] = s * p;d[l] = c * p;// Check for convergence.} while (Math.abs(e[l]) > eps*tst1);}d[l] = d[l] + f;e[l] = 0.0;}// Sort eigenvalues and corresponding vectors.for (int i = 0; i < n-1; i++) {int k = i;double p = d[i];for (int j = i+1; j < n; j++) {if (d[j] < p) {k = j;p = d[j];}}if (k != i) {d[k] = d[i];d[i] = p;for (int j = 0; j < n; j++) {p = V[j][i];V[j][i] = V[j][k];V[j][k] = p;}}}}// Nonsymmetric reduction to Hessenberg form.private void orthes () {// This is derived from the Algol procedures orthes and ortran,// by Martin and Wilkinson, Handbook for Auto. Comp.,// Vol.ii-Linear Algebra, and the corresponding// Fortran subroutines in EISPACK.int low = 0;int high = n-1;for (int m = low+1; m <= high-1; m++) {// Scale column.double scale = 0.0;for (int i = m; i <= high; i++) {scale = scale + Math.abs(H[i][m-1]);}if (scale != 0.0) {// Compute Householder transformation.double h = 0.0;for (int i = high; i >= m; i--) {ort[i] = H[i][m-1]/scale;h += ort[i] * ort[i];}double g = Math.sqrt(h);if (ort[m] > 0) {g = -g;}h = h - ort[m] * g;ort[m] = ort[m] - g;// Apply Householder similarity transformation// H = (I-u*u'/h)*H*(I-u*u')/h)for (int j = m; j < n; j++) {double f = 0.0;for (int i = high; i >= m; i--) {f += ort[i]*H[i][j];}f = f/h;for (int i = m; i <= high; i++) {H[i][j] -= f*ort[i];}}for (int i = 0; i <= high; i++) {double f = 0.0;for (int j = high; j >= m; j--) {f += ort[j]*H[i][j];}f = f/h;for (int j = m; j <= high; j++) {H[i][j] -= f*ort[j];}}ort[m] = scale*ort[m];H[m][m-1] = scale*g;}}// Accumulate transformations (Algol's ortran).for (int i = 0; i < n; i++) {for (int j = 0; j < n; j++) {V[i][j] = (i == j ? 1.0 : 0.0);}}for (int m = high-1; m >= low+1; m--) {if (H[m][m-1] != 0.0) {for (int i = m+1; i <= high; i++) {ort[i] = H[i][m-1];}for (int j = m; j <= high; j++) {double g = 0.0;for (int i = m; i <= high; i++) {g += ort[i] * V[i][j];}// Double division avoids possible underflowg = (g / ort[m]) / H[m][m-1];for (int i = m; i <= high; i++) {V[i][j] += g * ort[i];}}}}}// Complex scalar division.private transient double cdivr, cdivi;private void cdiv(double xr, double xi, double yr, double yi) {double r,d;if (Math.abs(yr) > Math.abs(yi)) {r = yi/yr;d = yr + r*yi;cdivr = (xr + r*xi)/d;cdivi = (xi - r*xr)/d;} else {r = yr/yi;d = yi + r*yr;cdivr = (r*xr + xi)/d;cdivi = (r*xi - xr)/d;}}// Nonsymmetric reduction from Hessenberg to real Schur form.private void hqr2 () {// This is derived from the Algol procedure hqr2,// by Martin and Wilkinson, Handbook for Auto. Comp.,// Vol.ii-Linear Algebra, and the corresponding// Fortran subroutine in EISPACK.// Initializeint nn = this.n;int n = nn-1;int low = 0;int high = nn-1;double eps = Math.pow(2.0,-52.0);double exshift = 0.0;double p=0,q=0,r=0,s=0,z=0,t,w,x,y;// Store roots isolated by balanc and compute matrix normdouble norm = 0.0;for (int i = 0; i < nn; i++) {if (i < low | i > high) {d[i] = H[i][i];e[i] = 0.0;}for (int j = Math.max(i-1,0); j < nn; j++) {norm = norm + Math.abs(H[i][j]);}}// Outer loop over eigenvalue indexint iter = 0;while (n >= low) {// Look for single small sub-diagonal elementint l = n;while (l > low) {s = Math.abs(H[l-1][l-1]) + Math.abs(H[l][l]);if (s == 0.0) {s = norm;}if (Math.abs(H[l][l-1]) < eps * s) {break;}l--;}// Check for convergence// One root foundif (l == n) {H[n][n] = H[n][n] + exshift;d[n] = H[n][n];e[n] = 0.0;n--;iter = 0;// Two roots found} else if (l == n-1) {w = H[n][n-1] * H[n-1][n];p = (H[n-1][n-1] - H[n][n]) / 2.0;q = p * p + w;z = Math.sqrt(Math.abs(q));H[n][n] = H[n][n] + exshift;H[n-1][n-1] = H[n-1][n-1] + exshift;x = H[n][n];// Real pairif (q >= 0) {if (p >= 0) {z = p + z;} else {z = p - z;}d[n-1] = x + z;d[n] = d[n-1];if (z != 0.0) {d[n] = x - w / z;}e[n-1] = 0.0;e[n] = 0.0;x = H[n][n-1];s = Math.abs(x) + Math.abs(z);p = x / s;q = z / s;r = Math.sqrt(p * p+q * q);p = p / r;q = q / r;// Row modificationfor (int j = n-1; j < nn; j++) {z = H[n-1][j];H[n-1][j] = q * z + p * H[n][j];H[n][j] = q * H[n][j] - p * z;}// Column modificationfor (int i = 0; i <= n; i++) {z = H[i][n-1];H[i][n-1] = q * z + p * H[i][n];H[i][n] = q * H[i][n] - p * z;}// Accumulate transformationsfor (int i = low; i <= high; i++) {z = V[i][n-1];V[i][n-1] = q * z + p * V[i][n];V[i][n] = q * V[i][n] - p * z;}// Complex pair} else {d[n-1] = x + p;d[n] = x + p;e[n-1] = z;e[n] = -z;}n = n - 2;iter = 0;// No convergence yet} else {// Form shiftx = H[n][n];y = 0.0;w = 0.0;if (l < n) {y = H[n-1][n-1];w = H[n][n-1] * H[n-1][n];}// Wilkinson's original ad hoc shiftif (iter == 10) {exshift += x;for (int i = low; i <= n; i++) {H[i][i] -= x;}s = Math.abs(H[n][n-1]) + Math.abs(H[n-1][n-2]);x = y = 0.75 * s;w = -0.4375 * s * s;}// MATLAB's new ad hoc shiftif (iter == 30) {s = (y - x) / 2.0;s = s * s + w;if (s > 0) {s = Math.sqrt(s);if (y < x) {s = -s;}s = x - w / ((y - x) / 2.0 + s);for (int i = low; i <= n; i++) {H[i][i] -= s;}exshift += s;x = y = w = 0.964;}}iter = iter + 1; // (Could check iteration count here.)// Look for two consecutive small sub-diagonal elementsint m = n-2;while (m >= l) {z = H[m][m];r = x - z;s = y - z;p = (r * s - w) / H[m+1][m] + H[m][m+1];q = H[m+1][m+1] - z - r - s;r = H[m+2][m+1];s = Math.abs(p) + Math.abs(q) + Math.abs(r);p = p / s;q = q / s;r = r / s;if (m == l) {break;}if (Math.abs(H[m][m-1]) * (Math.abs(q) + Math.abs(r)) <eps * (Math.abs(p) * (Math.abs(H[m-1][m-1]) + Math.abs(z) +Math.abs(H[m+1][m+1])))) {break;}m--;}for (int i = m+2; i <= n; i++) {H[i][i-2] = 0.0;if (i > m+2) {H[i][i-3] = 0.0;}}// Double QR step involving rows l:n and columns m:nfor (int k = m; k <= n-1; k++) {boolean notlast = (k != n-1);if (k != m) {p = H[k][k-1];q = H[k+1][k-1];r = (notlast ? H[k+2][k-1] : 0.0);x = Math.abs(p) + Math.abs(q) + Math.abs(r);if (x == 0.0) {continue;}p = p / x;q = q / x;r = r / x;}s = Math.sqrt(p * p + q * q + r * r);if (p < 0) {s = -s;}if (s != 0) {if (k != m) {H[k][k-1] = -s * x;} else if (l != m) {H[k][k-1] = -H[k][k-1];}p = p + s;x = p / s;y = q / s;z = r / s;q = q / p;r = r / p;// Row modificationfor (int j = k; j < nn; j++) {p = H[k][j] + q * H[k+1][j];if (notlast) {p = p + r * H[k+2][j];H[k+2][j] = H[k+2][j] - p * z;}H[k][j] = H[k][j] - p * x;H[k+1][j] = H[k+1][j] - p * y;}// Column modificationfor (int i = 0; i <= Math.min(n,k+3); i++) {p = x * H[i][k] + y * H[i][k+1];if (notlast) {p = p + z * H[i][k+2];H[i][k+2] = H[i][k+2] - p * r;}H[i][k] = H[i][k] - p;H[i][k+1] = H[i][k+1] - p * q;}// Accumulate transformationsfor (int i = low; i <= high; i++) {p = x * V[i][k] + y * V[i][k+1];if (notlast) {p = p + z * V[i][k+2];V[i][k+2] = V[i][k+2] - p * r;}V[i][k] = V[i][k] - p;V[i][k+1] = V[i][k+1] - p * q;}} // (s != 0)} // k loop} // check convergence} // while (n >= low)// Backsubstitute to find vectors of upper triangular formif (norm == 0.0) {return;}for (n = nn-1; n >= 0; n--) {p = d[n];q = e[n];// Real vectorif (q == 0) {int l = n;H[n][n] = 1.0;for (int i = n-1; i >= 0; i--) {w = H[i][i] - p;r = 0.0;for (int j = l; j <= n; j++) {r = r + H[i][j] * H[j][n];}if (e[i] < 0.0) {z = w;s = r;} else {l = i;if (e[i] == 0.0) {if (w != 0.0) {H[i][n] = -r / w;} else {H[i][n] = -r / (eps * norm);}// Solve real equations} else {x = H[i][i+1];y = H[i+1][i];q = (d[i] - p) * (d[i] - p) + e[i] * e[i];t = (x * s - z * r) / q;H[i][n] = t;if (Math.abs(x) > Math.abs(z)) {H[i+1][n] = (-r - w * t) / x;} else {H[i+1][n] = (-s - y * t) / z;}}// Overflow controlt = Math.abs(H[i][n]);if ((eps * t) * t > 1) {for (int j = i; j <= n; j++) {H[j][n] = H[j][n] / t;}}}}// Complex vector} else if (q < 0) {int l = n-1;// Last vector component imaginary so matrix is triangularif (Math.abs(H[n][n-1]) > Math.abs(H[n-1][n])) {H[n-1][n-1] = q / H[n][n-1];H[n-1][n] = -(H[n][n] - p) / H[n][n-1];} else {cdiv(0.0,-H[n-1][n],H[n-1][n-1]-p,q);H[n-1][n-1] = cdivr;H[n-1][n] = cdivi;}H[n][n-1] = 0.0;H[n][n] = 1.0;for (int i = n-2; i >= 0; i--) {double ra,sa,vr,vi;ra = 0.0;sa = 0.0;for (int j = l; j <= n; j++) {ra = ra + H[i][j] * H[j][n-1];sa = sa + H[i][j] * H[j][n];}w = H[i][i] - p;if (e[i] < 0.0) {z = w;r = ra;s = sa;} else {l = i;if (e[i] == 0) {cdiv(-ra,-sa,w,q);H[i][n-1] = cdivr;H[i][n] = cdivi;} else {// Solve complex equationsx = H[i][i+1];y = H[i+1][i];vr = (d[i] - p) * (d[i] - p) + e[i] * e[i] - q * q;vi = (d[i] - p) * 2.0 * q;if (vr == 0.0 & vi == 0.0) {vr = eps * norm * (Math.abs(w) + Math.abs(q) +Math.abs(x) + Math.abs(y) + Math.abs(z));}cdiv(x*r-z*ra+q*sa,x*s-z*sa-q*ra,vr,vi);H[i][n-1] = cdivr;H[i][n] = cdivi;if (Math.abs(x) > (Math.abs(z) + Math.abs(q))) {H[i+1][n-1] = (-ra - w * H[i][n-1] + q * H[i][n]) / x;H[i+1][n] = (-sa - w * H[i][n] - q * H[i][n-1]) / x;} else {cdiv(-r-y*H[i][n-1],-s-y*H[i][n],z,q);H[i+1][n-1] = cdivr;H[i+1][n] = cdivi;}}// Overflow controlt = Math.max(Math.abs(H[i][n-1]),Math.abs(H[i][n]));if ((eps * t) * t > 1) {for (int j = i; j <= n; j++) {H[j][n-1] = H[j][n-1] / t;H[j][n] = H[j][n] / t;}}}}}}// Vectors of isolated rootsfor (int i = 0; i < nn; i++) {if (i < low | i > high) {for (int j = i; j < nn; j++) {V[i][j] = H[i][j];}}}// Back transformation to get eigenvectors of original matrixfor (int j = nn-1; j >= low; j--) {for (int i = low; i <= high; i++) {z = 0.0;for (int k = low; k <= Math.min(j,high); k++) {z = z + V[i][k] * H[k][j];}V[i][j] = z;}}}/* ------------------------Constructor* ------------------------ *//** Check for symmetry, then construct the eigenvalue decompositionStructure to access D and V.@param Arg Square matrix*/public EigenvalueDecomposition (Matrix Arg) {double[][] A = Arg.getArray();n = Arg.getColumnDimension();V = new double[n][n];d = new double[n];e = new double[n];issymmetric = true;for (int j = 0; (j < n) & issymmetric; j++) {for (int i = 0; (i < n) & issymmetric; i++) {issymmetric = (A[i][j] == A[j][i]);}}if (issymmetric) {for (int i = 0; i < n; i++) {for (int j = 0; j < n; j++) {V[i][j] = A[i][j];}}// Tridiagonalize.tred2();// Diagonalize.tql2();} else {H = new double[n][n];ort = new double[n];for (int j = 0; j < n; j++) {for (int i = 0; i < n; i++) {H[i][j] = A[i][j];}}// Reduce to Hessenberg form.orthes();// Reduce Hessenberg to real Schur form.hqr2();}}/* ------------------------Public Methods* ------------------------ *//** Return the eigenvector matrix@return V*/public Matrix getV () {return new Matrix(V,n,n);}/** Return the real parts of the eigenvalues@return real(diag(D))*/public double[] getRealEigenvalues () {return d;}/** Return the imaginary parts of the eigenvalues@return imag(diag(D))*/public double[] getImagEigenvalues () {return e;}/** Return the block diagonal eigenvalue matrix@return D*/public Matrix getD () {Matrix X = new Matrix(n,n);double[][] D = X.getArray();for (int i = 0; i < n; i++) {for (int j = 0; j < n; j++) {D[i][j] = 0.0;}D[i][i] = d[i];if (e[i] > 0) {D[i][i+1] = e[i];} else if (e[i] < 0) {D[i][i-1] = e[i];}}return X;}private static final long serialVersionUID = 1; }

總結

以上是生活随笔為你收集整理的机器学习知识点(十九)矩阵特征值分解基础知识及Java实现的全部內容，希望文章能夠幫你解決所遇到的問題。

如果覺得生活随笔網站內容還不錯，歡迎將生活随笔推薦給好友。