Question

I am trying to do an efficient sparse matrix multiplication. Right now I am reading the data into memory and this is how my data structure looks like:

typedef struct node{
int x;
int y;
int value;
struct node* row;
struct node* col;
}node;

typedef struct matrix{
int height;
int width; 
node** rowList;
node** colList;
}matrix;

My current code for the insertion is:

void insert(matrix** M, int row_index, int col_index, int value)
{
    node* currNode=(node*)malloc(sizeof(node));
    currNode->x=row_index;
    currNode->y=col_index;
    currNode->value=value;

    if ((*M)->rowList[row_index] == NULL) { /* index is empty */
        currNode->row = NULL;
        (*M)->rowList[row_index] = currNode;
    }
    else if ((*M)->rowList[row_index]->y > col_index) { /* insert node to front */
        //printf("%d, %d\n", (*M)->rowList[row_index]->y, col_index);
        currNode->col = (*M)->rowList[row_index];
        (*M)->rowList[row_index] = currNode;
    }
    else if ((*M)->rowList[row_index]->y < col_index) { /* insert node to front */  
        node* rowptr = (node*)malloc(sizeof(node));
        rowptr = (*M)->rowList[row_index];
        while(rowptr->col!=NULL&&rowptr->col->y < col_index)
            rowptr=rowptr->col;

        currNode->col=rowptr->col;
        rowptr->col=currNode;
        //printf("-----------------%d\n", rowptr->y);
    }

    if ((*M)->colList[col_index] == NULL) { 
        currNode->col = NULL;
        (*M)->colList[col_index] = currNode;
    }
    else
    if ((*M)->colList[col_index]->x > row_index) { 
        //printf("%d, %d\n", (*M)->colList[col_index]->x, row_index);
        currNode->row = (*M)->colList[col_index];
        (*M)->colList[col_index] = currNode;
    }
}

In case of you ask, this is my print function:

void print_matrix(matrix *M){
    for(int i=0;i<M->height;i++){
        while(M->rowList[i]!=NULL){
            printf("i=%d, j=%d, v=%d\n",M->rowList[i]->x, M->rowList[i]->y,
                   M->rowList[i]->value);
            M->rowList[i]=M->rowList[i]->col;
        }
    }
}

For this input:

5,5
0,0,1
0,1,2
0,3,3
0,4,4

where (5,5) matrix dimensions and (0,0,1) = i,j,value, I get this:

i=0, j=0, v=1
i=0, j=1, v=2
i=0, j=3, v=3
i=0, j=4, v=4
i=0, j=4, v=4

For this input:

5,5
0,0,1
0,1,2
0,3,3
0,4,4
0,2,5

I get this:

i=0, j=0, v=1
i=0, j=1, v=2
i=0, j=2, v=5
i=0, j=2, v=5

I think the problem is here:

else if ((*M)->rowList[row_index]->y < col_index) {
    node* rowptr = (node*)malloc(sizeof(node));
    rowptr = (*M)->rowList[row_index];
    while(rowptr->col!=NULL&&rowptr->col->y < col_index)
        rowptr=rowptr->col;

        currNode->col=rowptr->col;
        rowptr->col=currNode;
    }
    [ ... ]

Somehow I remove one of the values when I add a new element that is smaller.

The question is : how can I get this code to load my sparse matrix values into memory using the data structure provided correctly?

Thank you ^^

Answer 1

Here:

if ((*M)->colList[col_index] == NULL) {
    currNode->col = NULL;
    (*M)->colList[col_index] = currNode;
}

where you write currNode->col, you should have written currNode->row. After making this change the output is correct for the second input file.

While looking at the code I noticed other odd things; for example, the print_matrix function also destroys the matrix ->col pointer chains. Also, in these two lines

    node* rowptr = (node*)malloc(sizeof(node));
    rowptr = (*M)->rowList[row_index];

you’re allocating memory and then immediately overwriting the pointer to it.