How to Count Number of Set Bits e.g. Number of 1's efficiently ?

This Question is asked in almost all core s/w companies & most of we know logarithmic solution of this problem but there exist a constant time solution using bit manipulations stuck !!!!:) See 2nd Method to solve the same problem in O(1)

1st Brian Kernighan’s Algorithm:
Subtraction of 1 from a number toggles all the bits (from right to left) till the rightmost set bit(including the righmost set bit). So if we subtract a number by 1 and do bitwise & with itself (n & (n-1)), we unset the righmost set bit. If we do n & (n-1) in a loop and count the no of times loop executes we get the set bit count.
Beauty of the this solution is number of times it loops is equal to the number of set bits in a given integer.

1 Initialize count: = 0
2 If integer n is not zero
(a) Do bitwise & with (n-1) and assign the value back to n
n: = n&(n-1)
(b) Increment count by 1
(c) go to step 2
3 Else return count

#include

/* Function to get no of set bits in binary
representation of passed binary no. */
int countSetBits(int n)
{
unsigned int count = 0;
while (n)
{
n &= (n-1) ;
count++;
}
return count;
}

/* Program to test function countSetBits */
int main()
{
int i = 16;
printf("%d", countSetBits(i));
getchar();
return 0;
}

Time Complexity O(logn)
Space Complexity O(1)
Run Here https://ideone.com/2y2KJ

2nd Method is Most Efficient One !!!!!!

Assuming that the integer is 32 bits, this is pretty good:

x = (x & 0x55555555) + ((x >> 1) & 0x55555555);
x = (x & 0x33333333) + ((x >> 2) & 0x33333333);
x = (x & 0x0F0F0F0F) + ((x >> 4) & 0x0F0F0F0F);
x = (x & 0x00FF00FF) + ((x >> 8) & 0x00FF00FF);
x = (x & 0x0000FFFF) + ((x >> 16) & 0x0000FFFF);

where
0x55555555 = 01010101 01010101 01010101 01010101
0x33333333 = 00110011 00110011 00110011 00110011
0x0F0F0F0F = 00001111 00001111 00001111 00001111
0x00FF00FF = 00000000 11111111 00000000 11111111
0x0000FFFF = 00000000 00000000 11111111 11111111


Notice that the hex constants are respectively alternate bits,
alternate pairs of bits, alternate groups of four bits, alternate
bytes, and the low-order half of the int.

The first statement determines the number of one-bits in each pair of
bits. The second statement adds adjacent pairs of bits to get the
number of bits in each group of four bits. Then these are added to get
the number of bits in each byte, short int, and finally in the whole
int.

but it works at low level ??

Suppose that the first four bits of x from the left are abcd. Lets separate the bits into pairs with a comma: ab,cd. The first four bits of the hex constant0x55... are 0101, or separated into pairs: 01,01. The logical product of x with this constant is 0b,0d. The first four bits of x>>1 are 0abc, or separated into pairs are 0a,bc. The logical product of this with the constant is 0a,0c. The sum 0b,0d + 0a,0c is a+b,c+d, where a+b = 00, 01, or 10, and b+c = 00, 01, or 10. Thus we have replaced each pair of bits in x with the sum of the two bits originally in the pair.

The next statement uses the constant 0x333.... The first four bits of this are 0011, split into pairs as 00,11. The logical product of the first four bits of x with this constant gives 00,c+d. Furthermore (a+b,c+d)>>2 = 00,a+b. Then 00,c+d + 00,a+b gives the four-bit quantity a+b+c+d, i.e., the number of one bits set in the first four bits of the original x.

The next statements continue to double the number of bits included in each sum.
& so on

#include
using namespace std;

int main()
{
int x=0x00000016;
x = (x & 0x55555555) + ((x >> 1) & 0x55555555);

cout<> 2) & 0x33333333);

cout<> 4) & 0x0F0F0F0F);

cout<> 8) & 0x00FF00FF);

cout<> 16) & 0x0000FFFF);

cout< return 0;

}

Time Complexity O(1)
Space Complexity O(1)
Run Here https://ideone.com/uDKGK
More Info http://graphics.stanford.edu/~seander/bithacks.html#CountBitsSetNaive