Hi all,

I'm mostly into my Java so I'm not even sure if I gave the post a proper heading. Basically I'm trying to read two matrices from the command line and print them out. Reading works as far as I can tell but when my program tries to call the printArr functions I get a segmentation fault. I think this is being caused because my pointers A and B are no longer pointing to the dynamic arrays once my function exits but the whole pointers thing is very new to me. Can anybody spot what the problem is?

#include <fstream>
#include <iostream>

using std::cout;
using std::endl;
using std::ofstream;

#include <math.h>
#include <stdlib.h>

typedef float** fArray;// in C++ multidimensional arrays are arrays of array pointers

void readArr(int, int, fArray);
void printArr(int, int, fArray);

int rows1;
int rows2;
int cols1;
int cols2;
int rowsum;
int colsum;
fArray A;
fArray B;
fArray Sum;

int main(int argc, char *argv[])
{
  if(argc != 5)// incorrect number of parameters
	{
	cout<< "Usage : [" << argv[0]<<" rows1 cols1 rows2 cols2] where rows and cols are the number of columns and rows in a given matrix."<<endl;
	} 

  else
	{
	
	  //atoi converts a string to an integer
	   rows1 = atoi(argv[1]);//number of rows matrix 1
	   cols1 = atoi(argv[2]);//number of cols matrix 1
	   rows2 = atoi(argv[3]);//number of rows matrix 2
	   cols2 = atoi(argv[4]);//number of cols matrix 2
  
	  if(cols1 == rows2)//we can multiply these matrices
		{
		//matrixMult(rows1, cols1, rows2, cols2);//call multiplication function with command line arguments
		readArr(rows1, cols1, A);
		readArr(rows2, cols2, B);
		printArr(rows1, cols1, A);
		printArr(rows2, cols2, B);
		
		}

		  else //must be unable to multiply matrices
		{
		cout<< "Unable to multiply matrices. The number of columns in matrix 1 must equal the number of rows in matrix 2. You supplied a value of" 			<<cols1<<" for the number of columns in matrix 1 and " <<rows2<<" for the number of rows in matrix 2."<<endl; 
		}
	}
}

void readArr(int rows, int cols, fArray array)
{
	array = new float*[rows];  //as noted above multidimensional arrays are arrays of arrays, new array[][] does not work so we need to create a 1d 
	for(int k = 0; k < rows; k++) // array and then iterate through the array with a for loop assigning each array as a further pointer which then 
	{                             //completes the "2d" array.
		array[k] = new float[cols];
	}
	for(int i = 0; i < rows; i++)
	{
		for (int j = 0; j < cols; j++)
		{
			std::cin >> array[i][j];
		}
	}
}

void printArr(int rows, int cols, fArray array)
{
	for(int i = 0; i < rows; i++)
	{
		cout<<endl;
		for(int j = 0; j < cols; j++)
		{
			cout<< array[i][j] <<" ";
		}	
	}
}

Recommended Answers

All 6 Replies

Just to mention the end goal of the program is to multiply 2 matrices, hence the currently unused ints and extra double ** declaration

You're correct, your array is going out of scope. You need to either pass your array by reference or by pointer. I suggest reference unless you explicitly need pointers.

void readArr(int rows, int cols, fArray& array)

And that should do it. You need to add the reference to both the declaration and the prototype.

commented: Clear, concise, excellent advice +2

Thanks for the reply, what do you mean by prototype? I assume declaration is the fArray A; etc.?

Ok I figured out that the prototype is the initial empty declaration of the function. Sorry about the lack of terminology knowledge but I'm coming from Java.

Everything works brilliantly now I just have a question or two about the mechanics.

& declares that I'm passing the address in memory, not the actual data from what I understand. Therefore using the & with the pointer is the equivalent of passing a java reference variable.

printArr still works without passing the reference but I'm wondering is this less efficient than passing in by reference? Does printArr(int rows, int cols, fArray array) create a local copy of the array to be manipulated, using more memory, and if so would adding the & operator to the prototype and function declaration for the printArr function result in less duplication and a more efficient program?

You're correct, print Arr is creating it's own copy. This obviously doesn't mean much with a small program, but passing by reference is good practice (unless you want the variable to fall out of scope at the end of the function). Adding the & to printArr is not necessary but is good practice.

That's excellent, thank you so much for the help. I was pulling my hair out trying to make sense of pointers, dereferencing and references but I think you've clarified nearly everything in 2 posts.

Be a part of the DaniWeb community

We're a friendly, industry-focused community of developers, IT pros, digital marketers, and technology enthusiasts meeting, networking, learning, and sharing knowledge.