Thread: 0 1 array? bit operation?

No, implemented as you describe, bitwise operators cannot help you. They work on integer types, not arrays of 1s and 0s.

You may be able to compact the array so that instead of storing 1s and 0s as presumably whole chars or ints, you could store them as bits that are part of ints or longs, or do similar to the collum and row vectors. Then you'd be able to use bitwise operators. If M or N are larger than 32(or possibly 64), this adds a good amount of complexity, and a fair amount if they aren't too. If the vectors are periodically read in full, this could pay off. But because of the complexity you should only do this if you have determined that accessing the vectors is a bottle neck.

There may be libraries that do this for you.

EDIT:
you can also implement a bit array using bitfeilds instead of bitwise operators, but it would take more code and would be slightly slower if the index is ever determined at runtime.

I thought that bit operators worked on any standard data type? So suppose we had a character you could do:

Code:

char a = 15;

for(int i = 0; i < 4; i++)
{
    printf("%d", (a >> i & 1));
}

That would print out for 1's right? I mean to ask it works with the raw bits regardless of the size. Is that not right?

It works on any integer type. But it will not work on an array as a whole.

Code:

#define BIT(a,n) (array)[(n)/COLS]&1<<((n)%CHAR_BIT)

Just woke up, but that's the general idea. Yes, you can manipulate the individual bits of the elements of an array. You may actually want different arguments there, especially if you switch to two-D. But the general idea is the same.

If you can flip the bits for a single instance of a given type, you can flip the bits of a single element of an array of any size of that same type.


Quzah.

When you say "giant" what do you mean? How sparse is the array (i.e. what is the ratio of 0's to 1's)? Having an array at all might not be the best way.