Faster uniform() in [0.0 - 1.0(

[Don]

tn wrote:
> tn Wrote:
> 
>> --------------------
>> int      []     271
>> int      [)     271
>> int      (]     283
>> int      ()     285
>>
>> long     []     372
>> long     [)     399
>> long     (]     401
>> long     ()     397
>>
>> float    []     286
>> float    [)     374
>> float    (]    5252
>> float    ()    5691
>>
>> double   []     348
>> double   [)     573
>> double   (]    5319
>> double   ()    5875
>>
>> real     []     434
>> real     [)     702
>> real     (]    2832
>> real     ()    3056
>> --------------------
>>
>> In my opinion floating point uniforms with (] or () as boundary types are unacceptably slow. I had to use 1 - uniform!"[)"(0.0, 1.0) instead of uniform!"(]"(0.0, 1.0) because of this issue. I would also expect versions using float and double to be faster than the version using real.
> 
> After further investigation it seems that the slowdown happens because of subnormal numbers in calculations. If there is an open boundary at zero then the call of nextafter in uniform returns a subnormal number. Perhaps next normal number could be used instead?

No, it just shouldn't convert (] into []. It should do [], and then 
check for an end point. Since the probability of actually generating a 
zero is 1e-4000, it shouldn't affect the speed at all <g>.