Thread: how to initialize a n-dimensional matrix and return it

Can anybody help me write a function like

Code:

func ( int n)
{

}

and inside the function I need to initialize an n-dimensional matrix, and return the same.

or if any one can refer me some book or links where i can learn to do the above it will be of great help to me.

start with a one dimensional array and go from there.

you're going to have to use loops and dynamic memory allocation.

http://www.cprogramming.com/tutorial/lesson3.html
http://www.cprogramming.com/tutorial/lesson6.html
http://www.cprogramming.com/tutorial/lesson8.html

FYI, it is generally better to use vector over dynamic arrays, but it doesn't sound like you have a choice in this assignment.

First of all its a self imposed assignment.

second thing I need to initialize n-dimensional matrix not array.

third I do give myself liberty to use array or vector, and I am expecting from myself a recursive solution and from you guys a little help

and so far no luck with my neurons !

By definition, a matrix is 2-dimensional. Presumably you're talking about an n x n matrix. Since you're free to do it any way you want, you should consider defining a Matrix class, and letting the dimensions m and n be data members (or just use one data member if you're only considering square matrices). You can handle allocation/deallocation in the constructor/destructor. If efficiency matters, you should use a 1-dimensional array or vector with m*n elements, and pretend it's two-dimensional by letting a[i][j] in the matrix correspond to i + j*n in the array/vector. This simplifies memory management and reduces memory fragmentation in case you want to have millions of matrices simultaneously.

> A matrix can therefore be n-dimensional where n exceeds 0 (a one dimensional matrix is a single value).
Wouldn't a single value correspond to n == 0, a row or column matrix to n == 1 (if the length of the row/column is > 1), and any other matrix to n == 2?

Originally Posted by robatino

By definition, a matrix is 2-dimensional. Presumably you're talking about an n x n matrix. Since you're free to do it any way you want, you should consider defining a Matrix class, and letting the dimensions m and n be data members (or just use one data member if you're only considering square matrices). You can handle allocation/deallocation in the constructor/destructor. If efficiency matters, you should use a 1-dimensional array or vector with m*n elements, and pretend it's two-dimensional by letting a[i][j] in the matrix correspond to i + j*n in the array/vector. This simplifies memory management and reduces memory fragmentation in case you want to have millions of matrices simultaneously.

Thanks for some explicit explanations, which i assumed the readers to know already.
But your explanation didnt help me to get anywhere ... I am still stuck !

I am talking about initializing n-dimensional matrix (square matrix), As far as i can see it can be done with vector or arrays.

I followed your idea of using a matrix class

Code:

class matrix
{
private :
int n;
public :
matrix(int i)
{

}
};

But I still dont know what to write inside the constructor so that i will have n-dimensional matrix (where each dimension each of length n). And there is a difference between millions of matrices and a matrix with million dimension.

Originally Posted by bawet

But I still dont know what to write inside the constructor so that i will have n-dimensional matrix (where each dimension each of length n). And there is a difference between millions of matrices and a matrix with million dimension.

That was answered in the second post in the thread.