HepDiagMatrix

HepDiagMatrix is a class for diagonal matrix. This is useful for a covariance matrix of measured quantities since they are uncorrelated to each other and therefore diagonal. It is obviously smaller and faster to manipulate.

Authors

Nobu Katayama and Mike Smyth

Declaration

#include "CLHEP/Matrix/DiagMatrix.h"

class HepDiagMatrix : public HepGenMatrix

Public Member Classes

HepDiagMatrix_row

Helper class to implement m[i][j]

Declaration: class HepDiagMatrix_row
Constructor: inline HepDiagMatrix_row(HepDiagMatrix&, int)
[]: inline HepDouble & operator[](int)

HepDiagMatrix_row_const

Helper class to implement m[i][j]

Declaration: class HepDiagMatrix_row_const
Constructor: inline HepDiagMatrix_row_const(const HepDiagMatrix&, int)
[]: inline const HepDouble & operator[](int) const

Public Member Functions

Constructor: inline HepDiagMatrix(); Default constructor. Gives 0 x 0 matrix.
Another HepDiagMatrix can be assigned to it.
Constructor: HepDiagMatrix(int p); HepDiagMatrix(int p, int); Gives p x p diagonal matrix.
With a second argument, either 0 or 1, the matrix is initialized.
0 means a zero matrix, 1 means the identity matrix.
Constructor: #ifdef HEP_USE_RANDOM; HepDiagMatrix(int p, HepRandom &r); #endif; Constructor with a Random object.
Copy constructor: HepDiagMatrix(const HepDiagMatrix &m1)
Destructor: virtual ~HepDiagMatrix()
=: HepDiagMatrix & operator = (const HepDiagMatrix &m2); Assignment operator.
To assign HepSymMatrix to HepDiagMatrix, use d << s
+=: HepDiagMatrix & operator += (const HepDiagMatrix &m2); Add a HepDiagMatrix.
-: HepDiagMatrix operator - () const; Unary minus, i.e. flip the sign of each element.
-=: HepDiagMatrix & operator -= (const HepDiagMatrix &m2); Subtract a HepDiagMatrix.
*=: HepDiagMatrix & operator *= (HepDouble t); Multiply a HepDiagMatrix by a floating number.
/=: HepDiagMatrix & operator /= (HepDouble t); Divide a HepDiagMatrix by a floating number.
(): HepDouble & operator()(int row, int col); const HepDouble & operator()(int row, int col) const; Read or write a matrix element. row must be equal to col.
** Note that indexing starts from (1,1) **
[]: inline HepDiagMatrix_row operator[] (int); inline HepDiagMatrix_row_const operator[] (int) const; Read or write a matrix element.
While it may not look like it, you simply do m[i][j] to get an element.
** Note that the indexing starts from [0][0] **
apply: HepDiagMatrix apply(HepDouble (*f)(HepDouble, int, int)) const; Apply a function to all elements of the matrix.
assign: void assign(const HepMatrix &m2); Assigns m2 to d, assuming m2 is a diagnal matrix.; void assign(const HepSymMatrix &m2); Assigns m2 to d, assuming m2 is a diagnal matrix.; void assign(const HepDiagMatrix &m2); Another form of assignment. For consistency.
determinant: HepDouble determinant() const; Calculate the determinant of the matrix.
fast: HepDouble & fast(int row, int col); const HepDouble & fast(int row, int col) const; Fast element access. Must be row >= col.
** Note that indexing starts from (1,1) **
inverse: HepDiagMatrix inverse(int &ierr) const; Invert a HepDiagMatrix. The matrix is not changed.
Returns 0 when successful, otherwise non-zero.
invert: void invert(int &ierr); Invert a HepDiagMatrix.
N.B. the contents of the matrix are replaced by the inverse.
Returns ierr = 0 when successful, otherwise non-zero.
This method has less overhead then inverse().
num_col: inline int num_col() const; Returns number of columns.
num_row: inline int num_row() const; Returns number of rows.
similarity: HepSymMatrix similarity(const HepMatrix &m1) const; Returns m1*s*m1.T(); HepDouble similarity(const HepVector &v) const; Returns v.T()*s*v (This is a scaler).
similarityT: HepSymMatrix similarityT(const HepMatrix &m1) const; Returns m1.T()*s*m1
sub: HepDiagMatrix sub(int min_row, int max_row) const; Returns a sub matrix of a HepDiagMatrix.; #ifdef HEP_CC_NEED_SUB_WITHOUT_CONST; HepDiagMatrix sub(int min_row, int max_row); #endif; SGI CC bug. I have to have both with/without const. I should not need one without const.; void sub(int row, const HepDiagMatrix &m1); Sub matrix of this HepDiagMatrix is replaced with m1.
T: HepDiagMatrix T() const; Returns the transpose of a HepDiagMatrix (which is itself).
trace: HepDouble trace() const; Calculate the trace of the matrix (sum of diagonal elements).

Non-Member Functions

+: HepMatrix operator + (const HepMatrix &m1, const HepDiagMatrix &d2); HepMatrix operator + (const HepDiagMatrix &d1, const HepMatrix &m2); HepDiagMatrix operator + (const HepDiagMatrix &m1, const HepDiagMatrix &d2); HepSymMatrix operator + (const HepSymMatrix &s1, const HepDiagMatrix &d2); HepSymMatrix operator + (const HepDiagMatrix &d1, const HepSymMatrix &s2); Addition operators.
-: HepMatrix operator - (const HepMatrix &m1, const HepDiagMatrix &d2); HepMatrix operator - (const HepDiagMatrix &d1, const HepMatrix &m2); HepDiagMatrix operator - (const HepDiagMatrix &d1, const HepDiagMatrix &d2); HepSymMatrix operator - (const HepSymMatrix &s1, const HepDiagMatrix &d2); HepSymMatrix operator - (const HepDiagMatrix &d1, const HepSymMatrix &s2); Subtraction operators.
*: HepMatrix operator * (const HepMatrix &m1, const HepDiagMatrix &m2); HepMatrix operator * (const HepDiagMatrix &m1, const HepMatrix &m2); Multiplication operators.
Note that m *= m1 is always faster than m = m * m1; HepDiagMatrix operator * (HepDouble t, const HepDiagMatrix &d1); HepDiagMatrix operator * (const HepDiagMatrix &d1, HepDouble t); Multiplication with a number.
/: HepDiagMatrix operator / (const HepDiagMatrix &m1, HepDouble t); Division with a number.
Note that d /= t is faster if you can use it.
<<: ostream & operator << (ostream &s, const HepDiagMatrix &q); Output to a stream.
dsum: HepDiagMatrix dsum(const HepDiagMatrix &d1, const HepDiagMatrix &d2); Direct sum of two diagonal matricies.

Example

CLHEP/test/testMatrix.cc

8 October 1997
EVC