Program Conversion Help

magda3227 · 06-11-2008, 05:48 PM

Hi. I am a college student who is doing research over the summer and one of my tasks is to convert a Wave Propagation Simulation code from MatLab into C. Needless to say, I have only very basic experience with C and I need a lot of help.

First off, I want to know if it is possible to write C code that will take the Kronecker Product of two matrices. I've only come across code that performs normal matrix multiplication thus far. If I can get assistance on that, it would be great.

Also, I'm having trouble with multidimensional arrays, which I obviously need.
I have 6 different cases (A-F), each that creates a different matrix of a different length/size than the others. I want to know how I am supposed to initialize my array before my switch statement so that I can store the values I want in it for each case. For example, in case A, the matrix needs to be a 2x1 matrix, but if case B were chosen, the matrix would need to be a 2x9 matrix.

I was also wondering about passing arrays to functions. I read a little bit about pointers, but I am not completely understanding how they work.

I know this is a lot of information/questions, but I hope someone will be able to assist me.

Thanks.

tabstop · 06-11-2008, 05:59 PM

If you can do it, you can program it. The Kronecker product doesn't look like it would be that bad.

The main problem is that matrices are not as fundamental to C as they are to Matlab. More to the point for you, a 2x1 matrix and a 2x9 matrix are not really compatible at all. You can perhaps build a struct that contains the dimensions and has a pointer to an array. In any rate, you're going to become an expert on memory management.

QuantumPete · 06-12-2008, 03:57 AM

Quote:

Originally Posted by magda3227

[whether] it is possible to write C code that [...]

Well, anything is possible really, you just have to have the time to implement it. To get the most out of this forum, ask specific questions. Start out with the most basic thing you want to do and show us what you have so far, we'll then be able to advise you on what to do.

QuantumPete

magda3227 · 06-12-2008, 03:36 PM

I have been able to make code that takes the Kronecker product of two matrices and store it within a new matrix. However, if the first matrix being multiplied has more rows or columns than the second matrix, the program gets stuck or returns incorrect values. I wanted to know if anyone can help me figure out how to fix this. Also, my coding is a lot of nested loops. I'm not very good at programming and feel that there is a less confusing/shorter code to accomplish the same thing. If there are any suggestions, please let me know.

Code:

#include <stdio.h>


/*Computes the Kronecker Product of two 2x2 matrices. If matrix A has more rows or columns than matrix B, the program does not finish. However, if matrix B has more rows or columns than A, the code works properly.*/

int main() 
{

	int cola, rowa, colb, rowb;
	cola = 2;
	rowa = 2;
	colb = 2;
	rowb = 2;

   int A[2][2] = {{1, 2}, {3, 4}},
       B[2][2] = {{0, 5},{6,7}};
	int C[rowa*rowb][cola*colb];

	
	
int i=0, j=0, k=0,l=0;

while(i<rowa-1)
{  
do{
	while(l<rowb-1)
{

	
  do{
	C[i+l][j+k] = A[i][j]*B[l][k];
	printf("%d\t",C[i+l][j+k]);
	k++;
	}while (k<colb);
	k=0; j++; 

   do{
	C[i+l][j+k+1] = A[i][j]*B[l][k];
	printf("%d\t",C[i+l][j+k+1]);
	k++;
	}while (k<colb);
	k=0; j=0; l++;

	printf("\n");



 do{
	C[i+l][j+k] = A[i][j]*B[l][k];
	printf("%d\t",C[i+l][j+k]);
	k++;
	}while (k<colb);
	 k=0; j++;

   do{
	C[i+l][j+k+1] = A[i][j]*B[l][k];
	printf("%d\t",C[i+l][j+k+1]);
	k++;
	}while (k<colb);
	k=0;  j=0;  i++; l=0; 

	printf("\n");

do{
	
	C[i+l+1][j+k] = A[i][j]*B[l][k];
	printf("%d\t",C[i+l+1][j+k]);
	k++;
	}while (k<colb);
	k=0; j++; 

	
   do{
	C[i+l+1][j+k+1] = A[i][j]*B[l][k];
	printf("%d\t",C[i+l+1][j+k+1]);
	k++;
	}while (k<colb);
	 k=0;  j=0; l++; 

	printf("\n");  

 do{
	C[i+l+1][j+k] = A[i][j]*B[l][k];
	printf("%d\t",C[i+l+1][j+k]);
	k++;
	}while (k<colb);
	k=0; j++; 

	
   do{

	C[i+l+1][j+k+1] = A[i][j]*B[l][k];
	printf("%d\t",C[i+l+1][j+k+1]);
	k++;
	}while (k<colb);
	k=0;

printf("\n");

}l=0; }while(j<cola-1);
} 


	/*printf("\n\n%d",C[0][3]);*/



return 0;
}

Thanks.

IceDane · 06-12-2008, 05:23 PM

When trying to come up with an algorithm for such a scenario, dividing it up into possible cases would help. Creating one algorithm that can handle dynamic sizes of (insert something here) is often hard, especially in this case by the looks of wikipedia.

With matrix A and B, the cases are:

1. A and B have same number of rows and columns.

2. Either one has more columns.

3. Either one has more rows.

4. Either one has more columns and rows.

5. Either one has more rows and the other has more columns.

Alright, five cases. Some might even be possible to migrate into the other cases/functions.

tabstop · 06-12-2008, 06:02 PM

You should be able to figure out a priori the size of C, which you have. The indices move in odometer style -- the second one moves faster, and when it resets the first one increments. And in each case, the first index goes around only once.

Hint: Consider a first matrix with dimension 3 and second matrix with dimension 5 (rows or columns, doesn't matter right now). That means the result should have dimension 15. Compute a/5 (integer division, so throw away remainder) and a%5 (here's the remainder) as a goes from 0 to 14. It starts 0,0; then 0,1; then 0,2; then 0,3; then 0,4; then 1,0; then 1,1; ...; then 2,4. This may give you some idea as to how to figure out indices.

magda3227 · 06-13-2008, 01:24 PM

I overcomplicated it. The code that works for any two matrices is the following:

Code:

#include <stdio.h>


/*Computes the Kronecker Product of two matrices and stores it as matrix C*/

int main() 
{

	int cola, rowa, colb, rowb;
	cola = 2;
	rowa = 3;
	colb = 3;
	rowb = 2;
      int a = rowa/rowb, b=rowa%rowb, c=cola/colb, d=cola%colb, e=1;


   int A[3][2] = {{1, 2}, {3, 4}, {1,0}},
       B[2][3] = {{0, 5, 2},{6,7, 3}};
	int C[rowa*rowb][cola*colb];

	int i,j,k,l;
	

	for(i=0; i<rowa; i++)
	{
		for(k=0; k<rowb; k++)
			{
				for(j=0; j<cola; j++)
					{
						for(l=0; l<colb; l++)
							{
								
								C[i+l+1][j+k+1] = A[i][j]*B[k][l];
								printf("%d\t",C[i+l+1][j+k+1]);

							}
					}printf("\n");
			} 
	}
	
	
	return 0;
}

Thanks to everyone who helped.

06-11-2008, 05:48 PM	#1
magda3227 Registered User Join Date: Jun 2008 Posts: 45	Program Conversion Help Hi. I am a college student who is doing research over the summer and one of my tasks is to convert a Wave Propagation Simulation code from MatLab into C. Needless to say, I have only very basic experience with C and I need a lot of help. First off, I want to know if it is possible to write C code that will take the Kronecker Product of two matrices. I've only come across code that performs normal matrix multiplication thus far. If I can get assistance on that, it would be great. Also, I'm having trouble with multidimensional arrays, which I obviously need. I have 6 different cases (A-F), each that creates a different matrix of a different length/size than the others. I want to know how I am supposed to initialize my array before my switch statement so that I can store the values I want in it for each case. For example, in case A, the matrix needs to be a 2x1 matrix, but if case B were chosen, the matrix would need to be a 2x9 matrix. I was also wondering about passing arrays to functions. I read a little bit about pointers, but I am not completely understanding how they work. I know this is a lot of information/questions, but I hope someone will be able to assist me. Thanks.
magda3227 is offline

06-11-2008, 05:59 PM	#2
tabstop and the Hat of Guessing Join Date: Nov 2007 Posts: 8,740	If you can do it, you can program it. The Kronecker product doesn't look like it would be that bad. The main problem is that matrices are not as fundamental to C as they are to Matlab. More to the point for you, a 2x1 matrix and a 2x9 matrix are not really compatible at all. You can perhaps build a struct that contains the dimensions and has a pointer to an array. In any rate, you're going to become an expert on memory management.
tabstop is offline

06-12-2008, 05:23 PM	#5
IceDane Ex scientia vera Join Date: Sep 2007 Posts: 426	When trying to come up with an algorithm for such a scenario, dividing it up into possible cases would help. Creating one algorithm that can handle dynamic sizes of (insert something here) is often hard, especially in this case by the looks of wikipedia. With matrix A and B, the cases are: 1. A and B have same number of rows and columns. 2. Either one has more columns. 3. Either one has more rows. 4. Either one has more columns and rows. 5. Either one has more rows and the other has more columns. Alright, five cases. Some might even be possible to migrate into the other cases/functions. __________________ "What's up, Doc?" "'Up' is a relative concept. It has no intrinsic value."
IceDane is offline

06-12-2008, 06:02 PM	#6
tabstop and the Hat of Guessing Join Date: Nov 2007 Posts: 8,740	You should be able to figure out a priori the size of C, which you have. The indices move in odometer style -- the second one moves faster, and when it resets the first one increments. And in each case, the first index goes around only once. Hint: Consider a first matrix with dimension 3 and second matrix with dimension 5 (rows or columns, doesn't matter right now). That means the result should have dimension 15. Compute a/5 (integer division, so throw away remainder) and a%5 (here's the remainder) as a goes from 0 to 14. It starts 0,0; then 0,1; then 0,2; then 0,3; then 0,4; then 1,0; then 1,1; ...; then 2,4. This may give you some idea as to how to figure out indices.
tabstop is offline

Thread Tools
Show Printable Version Email this Page
Display Modes
Linear Mode Switch to Hybrid Mode Switch to Threaded Mode

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Issue with program that's calling a function and has a loop	tigerfansince84	C++ Programming	9	11-12-2008 01:38 PM
Need help with my program...	Noah	C Programming	2	03-11-2006 07:49 PM
Decimal to Binary Conversion program	acidbeat311	C Programming	5	01-12-2006 10:26 PM
question about the loop in case conversion program	Elhaz	C++ Programming	8	09-20-2004 04:06 PM
My program, anyhelp	@licomb	C Programming	14	08-14-2001 10:04 PM