Embedded Systems Design - June 2008 - (Page 14)

programmer’s toolbox BY Jack W. Crenshaw Fleshing out the Matrix class T his month we continue with the process of developing a matrix class to accompany the vector class we developed earlier. As mathematical tools, vectors are a lot more useful when we can combine them with matrices. This is particularly true for the all-important matrix-vector product. In my last column, I gave you the skeleton of a matrix class. Now it’s time to put ﬂesh on the bones. From class Vector, I’ve borrowed the notion of using class variables to deﬁne the default dimensions of a matrix. This important step lets us adjust the size of objects created by the default constructor. This, in turn, lets us create arrays of matrices. The ﬁrst few additions to the class involve functions similar to those developed for class Vector, that let us examine and modify the class variables. These functions are: friend friend friend friend Jack Well, I suppose you don’t have to initialize the variables. You could always leave them unset, as long as you remember to set them before the ﬁrst matrix is constructed. However, as president and chairman of the board of the Society for the Elimination of Uninitialized Variables (SEUV), I don’t think much of this idea. Remember, these functions deal only with the dimensions of the default matrix—that is, a matrix created by the default constructor. Other matrices, created with explicit dimensions, or copied from somewhere else, are not affected by the class variables. Once an instance of class Matrix is created, its dimensions are ﬁxed. You can’t change gets to the meat of the them, and changes to the class Matrix class. variables don’t affect them either.You can, of course, ﬁnd out what the dimensions are. I’ve included two forms of the query functions: size_t size_t friend friend Rows(void); Cols(void); size_t Rows(const Matrix & a); size_t Cols(const Matrix & a); size_t GetMatrixRows(void); size_t GetMatrixCols(void); void SetMatrixRows(size_t m); void SetMatrixCols(size_t n); The usage should be pretty obvious. The downside of using class variables is that they must be explicitly initialized—an initializer in the class deﬁnition won’t do. What’s more, the initialization must appear at the global level, in your main program. The form is rather speciﬁc: size_t Matrix::matClassRows = 3; size_t Matrix::matClassCols = 3; This approach lets us choose the usage we prefer: m = A.Rows( ); or: m = Rows(A); Jack Crenshaw is a systems engineer and the author of Math Toolkit for Real-Time Programming. He holds a PhD in physics from Auburn University. E-mail him at jcrens@earthlink.net. UNDER CONSTRUCTION In my skeleton version of class Matrix, I only provided the default constructor. Any decent class needs at least a copy constructor, and we usually need others for constructing the object from other data types. In the case of matrices, the potential for different kinds of constructors is both rich and rather tricky, so I’ll take the time to discuss the choices in depth. 14 JUNE 2008 | embedded systems design | www.embedded.com http://www.embedded.com

Table of Contents Feed for the Digital Edition of Embedded Systems Design - June 2008

Embedded Systems Design - June 2008 Contents #Include Party Bit Programmer's Toolbox Cover Feature: Virtual Hardware Platforms for Embedded Software Validation Allocating Memory in MATLAB-to-C Code MDD and IDEs: Making the Twain Meet in Embedded Systems Design Avoid a Thrashing Guest Editor Advertising Index Break Points Marketplace

Embedded Systems Design - June 2008

For optimal viewing of this digital publication, please enable JavaScript and then refresh the page. If you would like to try to load the digital publication without using Flash Player detection, please click here.