AdHominem - 4 months ago 31

C Question

Title pretty much says it. I can only guess but I have no proof that this holds. In Python

`math.ceil(math.log(n) / math.log(2))`

Context: I'm creating a intbyte-to-char*binary function which only displays the necessary digits (id est no trailing zeros), thus I would need to dynamically allocate memory for the char array. However I assume there exists an upper bound which I could use to determine the length statically.

Answer

The formula is almost correct but you'll be out by one digit for exact powers of 2. The formula

`1 + math.floor(math.log(n) / math.log(2))`

would, I think, work better. But I wouldn't dwell on that too much simply because I'd never rely on a computer to compute `math.floor(math.log(n) / math.log(2))`

without the risk of the result being one less due to imprecision centred around floating point and the computation of a `log`

: for the latter you're at the mercy of your chipset.

Testing the length by repeated integer division by 2 until zero is attained would be more robust, and possibly faster. `log`

is not a cheap function computationally. You could even use the flashy bit fiddle `x & (x - 1)`

(Google it).