clockwerk::Matrix< R, C > Class Template Reference

Matrix math implementation. More...

#include <Matrix.hpp>

Public Member Functions
	Matrix ()
	Default constructor to initialize a matrix to all zeroes for ease of use.
	Matrix (floating_point elements)
	Constructor to initialize a matrix with all the same element.
	Matrix (const floating_point(&initial)[R][C])
	Matrix (const Matrix< R, C > &initial)
	Matrix (const std::array< std::array< floating_point, C >, R > &initial)
	Array constructor for the matrix class. Initializes from std::array.
	~Matrix ()
int16	str (char *output, size_t size) const
	Function to dump information on matrix.
int16	fromStr (const char *val)
	Set value of matrix from string in same format as str().
int16	set (uint32 row, uint32 col, const floating_point &value)
	Function to set a single value in the matrix.
int16	get (uint32 row, uint32 col, floating_point &result) const
	Function to get a single value in the matrix.
void	setFromArray (const floating_point *start_ptr, bool column_major=false)
	Function to set the values of the matrix row-wise.
void	getAsArray (floating_point *start_ptr, bool column_major=false) const
	Function to get the values of the matrix row-wise.
floating_point	get (uint32 row, uint32 col) const
	Function to get a single value in the matrix.
void	getCopy (Matrix< R, C > &result) const
	Function to get a copy of the matrix.
Matrix< R, C > &	operator= (const Matrix< R, C > &other)
	Equals operator overload for matrix.
floating_point *	operator[] (uint32 idx)
	Function to return a matrix row or vector value.
std::pair< uint32, uint32 >	size () const
	Function to get the size of the matrix.
void	max (floating_point &result, std::pair< uint32, uint32 > &index) const
	Function to return the maximum value in the matrix.
void	min (floating_point &result, std::pair< uint32, uint32 > &index) const
	Function to return the minimum value in the matrix.
int16	det (floating_point &result) const
	Function to return the determinant of the matrix.
int16	chol (Matrix< R, C > &retval) const
	Take the cholesky decomposition of this matrix.
Matrix< R, C >	chol () const
int16	inverse (Matrix< R, C > &result) const
	Function to return the inverse of the matrix.
Matrix< R, C >	inverse () const
int16	pseudoinverse (Matrix< C, R > &result) const
Matrix< C, R >	pseudoinverse () const
void	transpose (Matrix< C, R > &result) const
	Function to return the transpose of the matrix.
Matrix< C, R >	transpose () const
int16	trace (floating_point &result) const
	Function to return the trace of the matrix.
void	setToZeros ()
	Function to set all elements of the matrix to zero.
int16	identity ()
	Function to set matrix to identity, if it is a square matrix.
int16	eye ()

Public Attributes
std::array< std::array< floating_point, C >, R >	values

Protected Member Functions
int16	_checkLookupBoundaries (uint32 start_r, uint32 end_r, uint32 start_c, uint32 end_c) const
int16	_LUPDecompose (floating_point *A[R], uint32 P[R+1]) const
	Function to take a 2-d matrix represented by A and decompose it into LU form.

Detailed Description

template<uint32 R, uint32 C>
class clockwerk::Matrix< R, C >

Matrix math implementation.

This file defines a base class for computational matrix math. It is designed for use in real-time, embedded mode or with simulation. As such, it contains certain limitations on dynamic allocation to make it real-time compliant.

Note: The naming convention observed for matrix math is as follows:

• Capital letters represent matrices
• Lowercase letters represent scalars
• All vectors are 1-dimensional matrices for the purpose of math
• Pass by reference return values are all named "result"
• The letter used in math indicates the order of operations - A/a comes first, then B/b. A/a or B/b always represent the "other" value in an operation

Important: The matrix class is templated such that it can be used with a generic type floating_point, but should only be used on native types, as other implementations could result in undefined behavior.

Constructor & Destructor Documentation

Matrix() [1/5]

template<uint32 R, uint32 C>

clockwerk::Matrix< R, C >::Matrix

(

)

Default constructor to initialize a matrix to all zeroes for ease of use.

Matrix() [2/5]

template<uint32 R, uint32 C>

clockwerk::Matrix< R, C >::Matrix

(

floating_point

elements

)

Constructor to initialize a matrix with all the same element.

Matrix() [3/5]

template<uint32 R, uint32 C>

clockwerk::Matrix< R, C >::Matrix

(

const floating_point(&)

initial[R][C]

)

Constructor for Matrix class with initialization. Initializes matrix to values passed in via array

Matrix() [4/5]

template<uint32 R, uint32 C>

clockwerk::Matrix< R, C >::Matrix

(

const Matrix< R, C > &

initial

)

Copy constructor for Matrix class. Copies data from matrix object to the current instance

Matrix() [5/5]

template<uint32 R, uint32 C>

clockwerk::Matrix< R, C >::Matrix

(

const std::array< std::array< floating_point, C >, R > &

initial

)

Array constructor for the matrix class. Initializes from std::array.

~Matrix()

template<uint32 R, uint32 C>

clockwerk::Matrix< R, C >::~Matrix ( )

inline

Destructor – doesn't do anything because we don't dynamically allocate

Member Function Documentation

_checkLookupBoundaries()

template<uint32 R, uint32 C>

int16 clockwerk::Matrix< R, C >::_checkLookupBoundaries ( uint32 start_r, uint32 end_r, uint32 start_c, uint32 end_c ) const

protected

Function to check, given a set of submatrix boundaries, that those boundaries are valid for the current matrix

_LUPDecompose()

template<uint32 R, uint32 C>

int16 clockwerk::Matrix< R, C >::_LUPDecompose ( floating_point * A[R], uint32 P[R+1] ) const

protected

Function to take a 2-d matrix represented by A and decompose it into LU form.

Parameters

A	A 2-d matrix represented as a double pointer After ops A contains a copy of both matrices L-E and U as A=(L-E)+U such that P*A=L*U.
P	The permutation matrix to hold info on matrix changes

Returns

Error code corresponding to codes in clockwerkerrors.h

Note

Code borrowed/updated from sample C code on wikipedia here: https://en.wikipedia.org/wiki/LU_decomposition

chol() [1/2]

template<uint32 R, uint32 C>

Matrix< R, C > clockwerk::Matrix< R, C >::chol ( ) const

inline

chol() [2/2]

template<uint32 R, uint32 C>

int16 clockwerk::Matrix< R, C >::chol

(

Matrix< R, C > &

retval

)

const

Take the cholesky decomposition of this matrix.

Parameters

retval

Implicit return of calculation result

Returns

Error code corresponding to success/failure

Note

Must be positive semidefinite matrix! Use this if you know what you're doing.

Because of this the direct return is super dangerous. Be very careful.

det()

template<uint32 R, uint32 C>

int16 clockwerk::Matrix< R, C >::det

(

floating_point &

result

)

const

Function to return the determinant of the matrix.

---

Operations to return important matrix information

Parameters

result

PBR return of determinant

Returns

Integer value corresponding to error codes in clockwerkerrrors.h

eye()

template<uint32 R, uint32 C>

int16 clockwerk::Matrix< R, C >::eye ( )

inline

fromStr()

template<uint32 R, uint32 C>

int16 clockwerk::Matrix< R, C >::fromStr

(

const char *

val

)

Set value of matrix from string in same format as str().

Parameters

val	The value to set matrix from of the form [[1,2][3,4][5,6]]

Returns

Error code corresponding to success/failure

get() [1/2]

template<uint32 R, uint32 C>

floating_point clockwerk::Matrix< R, C >::get	(	uint32	row,
		uint32	col ) const

Function to get a single value in the matrix.

Parameters

row	The row index
col	The column index

Returns

Matrix value at index

Note

Does not bounds check and is unsafe for embedded operations

get() [2/2]

template<uint32 R, uint32 C>

int16 clockwerk::Matrix< R, C >::get	(	uint32	row,
		uint32	col,
		floating_point &	result ) const

Function to get a single value in the matrix.

Parameters

row	The row index
col	The column index
value	PBR return of the value in the matrix

Returns

Integer value corresponding to error codes in clockwerkerrrors.h

getAsArray()

template<uint32 R, uint32 C>

void clockwerk::Matrix< R, C >::getAsArray	(	floating_point *	start_ptr,
		bool	column_major = false ) const

Function to get the values of the matrix row-wise.

Parameters

start_ptr	The data address at which write should begin.
column_major	Flag indicating whether matrix should output column major. Default is false

Note

start_ptr must represent an array or similar type of at least R*C values

Not protected. May segfault if not used properly

The matrix [1 2; 3 4] would output as an array [1 2 3 4]

getCopy()

template<uint32 R, uint32 C>

void clockwerk::Matrix< R, C >::getCopy

(

Matrix< R, C > &

result

)

const

Function to get a copy of the matrix.

Parameters

result

PBR return of a copy of this matrix

identity()

template<uint32 R, uint32 C>

int16 clockwerk::Matrix< R, C >::identity

(

)

Function to set matrix to identity, if it is a square matrix.

Returns

Error code corresponding to success/failure

inverse() [1/2]

template<uint32 R, uint32 C>

Matrix< R, C > clockwerk::Matrix< R, C >::inverse ( ) const

inline

inverse() [2/2]

template<uint32 R, uint32 C>

int16 clockwerk::Matrix< R, C >::inverse

(

Matrix< R, C > &

result

)

const

Function to return the inverse of the matrix.

Parameters

result

PBR return of matrix inverse

Returns

Integer value corresponding to error codes in clockwerkerrrors.h

max()

template<uint32 R, uint32 C>

void clockwerk::Matrix< R, C >::max	(	floating_point &	result,
		std::pair< uint32, uint32 > &	index ) const

Function to return the maximum value in the matrix.

Parameters

result

PBR return of maximum value

min()

template<uint32 R, uint32 C>

void clockwerk::Matrix< R, C >::min	(	floating_point &	result,
		std::pair< uint32, uint32 > &	index ) const

Function to return the minimum value in the matrix.

Parameters

result

PBR return of minimum value

operator=()

template<uint32 R, uint32 C>

Matrix< R, C > & clockwerk::Matrix< R, C >::operator=

(

const Matrix< R, C > &

other

)

Equals operator overload for matrix.

operator[]()

template<uint32 R, uint32 C>

floating_point * clockwerk::Matrix< R, C >::operator[]

(

uint32

idx

)

Function to return a matrix row or vector value.

Parameters

idx	The row index to return

Returns

A pointer to the row element of the matrix. Nullptr if out of range

pseudoinverse() [1/2]

template<uint32 R, uint32 C>

Matrix< C, R > clockwerk::Matrix< R, C >::pseudoinverse ( ) const

inline

pseudoinverse() [2/2]

template<uint32 R, uint32 C>

int16 clockwerk::Matrix< R, C >::pseudoinverse

(

Matrix< C, R > &

result

)

const

set()

template<uint32 R, uint32 C>

int16 clockwerk::Matrix< R, C >::set	(	uint32	row,
		uint32	col,
		const floating_point &	value )

Function to set a single value in the matrix.

---

Getters and setters

Parameters

row	The row index
col	The column index
value	The value to set the matrix to

Returns

Integer value corresponding to error codes in clockwerkerrrors.h

setFromArray()

template<uint32 R, uint32 C>

void clockwerk::Matrix< R, C >::setFromArray	(	const floating_point *	start_ptr,
		bool	column_major = false )

Function to set the values of the matrix row-wise.

Parameters

start_ptr	The data address at which read should begin.
column_major	Flag indicating whether matrix should input column major. Default is false

Note

start_ptr must represent an array or similar type of at least R*C values

Not protected. May segfault if not used properly

The array [1 2 3 4] would read into a matrix as [1 2; 3 4]

setToZeros()

template<uint32 R, uint32 C>

void clockwerk::Matrix< R, C >::setToZeros

(

)

Function to set all elements of the matrix to zero.

size()

template<uint32 R, uint32 C>

std::pair< uint32, uint32 > clockwerk::Matrix< R, C >::size ( ) const

inline

Function to get the size of the matrix.

---

Operations to get matrix info

str()

template<uint32 R, uint32 C>

int16 clockwerk::Matrix< R, C >::str	(	char *	output,
		size_t	size ) const

Function to dump information on matrix.

Function to dump information on matrix to string

trace()

template<uint32 R, uint32 C>

int16 clockwerk::Matrix< R, C >::trace

(

floating_point &

result

)

const

Function to return the trace of the matrix.

Parameters

result

PBR return of matrix trace

Returns

Integer value corresponding to error codes in clockwerkerrrors.h

transpose() [1/2]

template<uint32 R, uint32 C>

Matrix< C, R > clockwerk::Matrix< R, C >::transpose

(

)

const

transpose() [2/2]

template<uint32 R, uint32 C>

void clockwerk::Matrix< R, C >::transpose

(

Matrix< C, R > &

result

)

const

Function to return the transpose of the matrix.

Parameters

result

PBR return of matrix transpose

Note

Overloaded for return as well

Member Data Documentation

values

template<uint32 R, uint32 C>

std::array<std::array<floating_point, C>, R> clockwerk::Matrix< R, C >::values

The actual values held by the matrix – a two dimensional array of values indexed as (row, column). NOTE: Public for ease of access and speed (no need to use setter/getter), and functions in Matrix and Safemath libraries are tested safe. Behavior using matrices outside of clockwerk libraries should use setter/getter, rather than direct access, for safety.

---

The documentation for this class was generated from the following file:

• /Users/mickey/Documents/Projects/warptwin/warpos/clockwerk/src/core/Matrix.hpp