bearophile wrote:
Andrei Alexandrescu:
3. How should the global rng be initialized?
Automatically seeded with the time at the beginning of the program. Of course
the seed can be set again in any moment.
4. While we're at it, should uniform(a, b) generate by default something
in [a, b] or [a, b)? Someone once explained to me that generating [a, b]
for floating point numbers is the source of all evils and that Hitler,
Stalin and Kim Il Sung (should he still be alive) must be using that
kind of generator. Conversely, generating [a, b) is guaranteed to bring
in the long term everlasting peace to Earth. My problem however is that
in the integer realm I always want to generate [a, b]. Furthermore, I
wouldn't be happy if the shape of the interval was different for
integers and floating point numbers. How to break this conundrum? Don't
forget that we're only worrying about defaults, explicit generation is
always possible with self-explanatory code:
auto rng = Random(unpredictableSeed);
auto a = 0.0, b = 1.0;
auto x1 = uniform!("[]")(rng, a, b);
auto x2 = uniform!("[)")(rng, a, b);
auto x3 = uniform!("(]")(rng, a, b);
auto x4 = uniform!("()")(rng, a, b);
That's awful and ugly. My suggestions are simple (copied from my dlibs and the
random std lib of Python):
random() => floating point [0, 1)
randInt(a=0, b) => integral [a, b]
randRange(a=0, b) => integral [a, b)
uniform(a, b) => floating point [a, b)
normal(a, b) => good quality normally-distributed number with given std dev and
avg.
Leaving normal() aside (does your implementation implement the ziggurat
algorithm? If so, would you agree to contribute it to phobos?), I don't
see how having to memorize four names instead of one is necessarily
awesome and beautiful. Besides uniform() renders "random" redundant.
Hardly a pinnacle of good API design, see realloc() which essentially
makes malloc() and free() redundant.
Andrei
