Proposal: Move 'split' into a separate class in System.Random

Bryan O'Sullivan bos at serpentine.com
Thu Sep 16 12:48:12 EDT 2010


On Wed, Sep 15, 2010 at 12:57 PM, Isaac Dupree <
ml at isaac.cedarswampstudios.org> wrote:

>
> Citation/argument for "It is sure to introduce correlations between the
> forked generators"?  Did someone prove that for all possible PRNGs?  Or are
> you just claiming a lack of anyone having proved it to be sound?
>

Well, it's necessarily the case that if the children of a split depend only
on the state of their parent, their states must be correlated. For them to
be uncorrelated, one would have to acquire its state from another source.

Split on StdGen worked fine for me displaying random smoke.  The only effect
> of the random numbers here is how it looks.  And it looked fine.  Actually I
> re-implemented that with a linear congruential generator in C for speed.
>  Just saying, *some* uses of random numbers don't need them to have very
> good properties.
>

That's true, but that's also not an argument in favour of the PRNG being bad
:-)
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.haskell.org/pipermail/libraries/attachments/20100916/d87ba5e6/attachment.html


More information about the Libraries mailing list