carvalho.miguel@[EMAIL PROTECTED]
wrote:
> On 27 Abr, 21:37, Krzysztof Mitko <kmi...@[EMAIL PROTECTED]
> wrote:
>> carvalho.mig...@[EMAIL PROTECTED]
twierdzi, że:
>>
>>> im trying to generate a double precision random number between 0 and
>>> 1, but i keep getting 0 only.
>>> static double randomDouble()
>>> {
>>> return static_cast<double>(rand() / RAND_MAX);
>>> }
>>
>> RAND_MAX is an integer
>>
>> rand() returns an integer
>>
>> (rand() / RAND_MAX) is a division of integer by integer, suprisingly
>> giving the integer :). Further cast doesn't matter, since you can
>> only get 0 or 1 (if rand() results in RAND_MAX value).
>> One of them should be a double. My manual says that the proper
>> formula is:
>>
>> return rand() / (RAND_MAX + 1.0);
>>
>> --
>> If an experiment works, something has gone wrong.
>
> nice, it worked. and you're right i was dividing 2 integers getting a
> new integer, really nubish of me.
> thanks for the help and explanation =)
Alternative is:
return static_cast<doube>( rand() ) / RAND_MAX;
This casts the return from rand() to a double, then devides by RAND_MAX.
Since at this pont rand() is a double, you will get floating point
divisiion.
--
Jim Langston
tazmaster@[EMAIL PROTECTED]
|
