Thread: linked list vol.2

Because you didn't tell main() to actually use the value returned by Function().

Try this in main():

Code:

anchor = Function(anchor);

EDIT: the only other way to get it work as intended is to change to function's arguments in order to simulate call-by-reference or use global variables (ugly).

Greets,
Philip

Are you sure that you are supposed to create a completely new list, and not just modify the existing list?

--
Mats

Originally Posted by rabers

I didnt say I have to create new list.Its just that I thought I can do all the modification with another list and pointing it with the pointer sent as an argument (head)

But since it is just a pointer to the struct, you can not change what the head/anchor value is - you can only change the contents of what head points to, not head itself. To change the value of anchor in main, it would have to be a pointer to pointer.

So, if this is supposed to work, you should not recreate the list, but rather modify the existing one - at least, that's the best understanding I can see.

--
Mats

the problem is:
its a matrix (each struct point to 2 other structs- right and down). its a sparse matrix and the function should turn it into the transposed matrix.
thats why I created a new one and placed the original items at a "transposed way" at the new one and freed the original

Originally Posted by rabers

the problem is:
its the a matrix (each struct point to 2 other structs- right and down). its a sparse matrix and the function should turn it into the transposed matrix.
thats why I created a new one and placed the original items at a "transposed way" at the new one and freed the original

Ok, so I don't know if that's hard or easy to do without creating a new matrix, but I suspect you could do it the other way around: Create a copy, then put the items back in the original matrix, in the transposed order. If I understand matrix transposition correctly, all you should need to do is replace all down-pointers with right-pointers, and all right-pointers with down-pointers, respectively.

--
Mats

its a sparse matrix and the function should turn it into the transposed matrix.

You can easily do inplace transposition of a matrix, whether it's sparse or not:

Code:

for(i = 0; i < ROWS; i++)
        for(j = 0; j < COLS; j++)
                A[i][j] = A[j][i];

Simply substitute array subscripts with the appropriate set-/get-functions, maybe something along the lines of

Code:

                matrix_set(A, i, j, matrix_get(A, j, i));

If efficiency is a concern, don't use linked data structures for storing matrices.

Greets,
Philip

EDIT: now that I posted the question, I can see the problem. For matrices where ROWS != COLS, this doesn't work as easily. Let me start a new post...

Ok, so my ultra-clever approach doesn't work with MxN-matrices where M != N, so let's see...

Aside from the fact that it's pretty stupid to return something and then not making use of it (which would solve all your problems immediately), try this:

In Function(), create a new matrix A^T incrementally and make it the transposed of your original matrix A. Then take the first entry of A^T, i.e. A^T[0][0] in array notation, and substitute it with the corresponding node from A. Then give this node its correct value, including adjusting its pointers, and free A^T[0][0] and the remaining part of the original matrix A.

This way, the pointer in main() points to the same address, which then is the transposed matrix.

Greets,
Philip

Originally Posted by Snafuist

Ok, so my ultra-clever approach doesn't work with MxN-matrices where M != N, so let's see...

Aside from the fact that it's pretty stupid to return something and then not making use of it (which would solve all your problems immediately), try this:

In Function(), create a new matrix A^T incrementally and make it the transposed of your original matrix A. Then take the first entry of A^T, i.e. A^T[0][0] in array notation, and substitute it with the corresponding node from A. Then give this node its correct value, including adjusting its pointers, and free A^T[0][0] and the remaining part of the original matrix A.

This way, the pointer in main() points to the same address, which then is the transposed matrix.

Greets,
Philip

I still think that swapping all down & right links will do the job.

--
Mats

Originally Posted by rabers

thanx. I just saw it now.
as I said earlyer - I fixed it and it happen to be your way
thanx again

I _STILL_ think that the correct way is to swap the right/down pointers. Sorry to repeat myself.

--
Mats