numpy performance and random numbers

[David Cournapeau]

On Sun, Dec 20, 2009 at 6:47 PM, Lie Ryan <lie.1296 at gmail.com> wrote:
> On 12/20/2009 2:53 PM, sturlamolden wrote:
>>
>> On 20 Des, 01:46, Lie Ryan<lie.1... at gmail.com>  wrote:
>>
>>> Not necessarily, you only need to be certain that the two streams don't
>>> overlap in any reasonable amount of time. For that purpose, you can use
>>> a PRNG that have extremely high period like Mersenne Twister and puts
>>> the generators to very distant states.
>>
>> Except there is no way to find two very distant states and prove they
>> are distant enough.
>>
> Except only theoretical scientist feel the need to prove it and perhaps
> perhaps for cryptographic-level security. Random number for games, random
> number for tmp files, and 99.99% random number users doesn't really need
> such proves.

But the OP case mostly like falls in your estimated 0.01% case. PRNG
quality is essential for reliable Monte Carlo procedures. I don't
think long period is enough to guarantee those good properties for //
random generators - at least it is not obvious to me.

David