Number of 1 Bits

Signal

"Count set bits" with no better-than-32-steps requirement is fine to brute force bit-by-bit, but the sharper tool is peeling off the lowest set bit directly — a trick worth having automatic, since it reappears any time a problem cares about "how many 1-bits" rather than "which bits."

Approach

n & (n - 1) clears exactly the lowest set bit of n (subtracting 1 flips that bit and every trailing zero below it, and ANDing with the original wipes just that bit). Repeat until n is 0, counting how many times you did it — that count is the number of set bits, and you only loop once per set bit rather than once per bit position.

Skeleton

count = 0
while n:
    n &= n - 1
    count += 1
return count

Solution

def hamming_weight(n: int) -> int:
    count = 0
    while n:
        n &= n - 1
        count += 1
    return count

Complexity

O(k) time where k is the number of set bits (at most 32) — strictly faster than checking all 32 positions when the number is sparse. O(1) space.

Pitfalls

• Shifting and checking all 32 bit positions (for i in range(32): n >> i & 1) is also correct and a fine first answer, but it's a fixed 32 iterations regardless of how many bits are actually set — mention n & (n-1) as the tighter version if asked to optimize.
• Python's bin(n).count('1') gets the right answer in one line and is perfectly reasonable to mention, but it's worth being able to derive the bit-trick by hand since that's usually the actual ask.