3 ms·
The article claims the scalar exponential solution is O(1) - this is incorrect because the exponential of a scalar is still O(logN).
by haykmartiros 8y ago
The article claims the scalar exponential solution is O(1) - this is incorrect because the exponential of a scalar is still O(logN).
- contravariant 8y agoWhat N? Floating point operations might be limited in precision (and range to some extend), but pretty much any you'll normally encounter will be O(1). Unless you require arbitrary precision.
- umanwizard 8y agoBy exactly the same reasoning, every algorithm is O(1) because your computer is a finite object and so the size of the data is bounded.