Hi Duncan--
Nice simulation!
The absolute difference in probabilities is small, but the maximum
relative difference grows from something negligible to almost 2 as m
approaches 2**31.
Because the L_1 distance between the uniform distribution on {1, ..., m}
and what you actually get is large, there have to be test functions whose
expectations are quite different under the two distributions. Whether those
correspond to commonly used statistics or not, I have no idea.
Regarding backwards compatibility: as a user, I'd rather the default
sample() do the best possible thing, and take an extra step to use
something like sample(..., legacy=TRUE) if I want to reproduce old results.
Regards,
Philip
On Wed, Sep 19, 2018 at 9:50 AM Duncan Murdoch <murdoch.duncan at gmail.com>
wrote:
On 19/09/2018 12:23 PM, Philip B. Stark wrote:
No, the 2nd call only happens when m > 2**31. Here's the code:
Yes, you're right. Sorry!
So the ratio really does come close to 2. However, the difference in
probabilities between outcomes is still at most 2^-32 when m is less
than that cutoff. That's not feasible to detect; the only detectable
difference would happen if some event was constructed to hold an
abundance of outcomes with especially low (or especially high)
probability.
As I said in my original post, it's probably not hard to construct such
a thing, but as I've said more recently, it probably wouldn't happen by
chance. Here's one attempt to do it:
Call the values from unif_rand() "the unif_rand() outcomes". Call the
values from sample() the sample outcomes.
It would be easiest to see the error if half of the sample() outcomes
used two unif_rand() outcomes, and half used just one. That would mean
m should be (2/3) * 2^32, but that's too big and would trigger the other
version.
So how about half use 2 unif_rands(), and half use 3? That means m =
(2/5) * 2^32 = 1717986918. A good guess is that sample() outcomes would
alternate between the two possibilities, so our event could be even
versus odd outcomes.
Let's try it:
> m <- (2/5)*2^32
> m > 2^31
> x <- sample(m, 1000000, replace = TRUE)
> table(x %% 2)
0 1
399850 600150
Since m is an even number, the true proportions of evens and odds should
be exactly 0.5. That's some pretty strong evidence of the bug in the
generator. (Note that the ratio of the observed probabilities is about
1.5, so I may not be the first person to have done this.)
I'm still not convinced that there has ever been a simulation run with
detectable bias compared to Monte Carlo error unless it (like this one)
was designed specifically to show the problem.
Duncan Murdoch
(RNG.c, lines 793ff)
double R_unif_index(double dn)
{
double cut = INT_MAX;
switch(RNG_kind) {
case KNUTH_TAOCP:
case USER_UNIF:
case KNUTH_TAOCP2:
cut = 33554431.0; /* 2^25 - 1 */
break;
default:
break;
}
double u = dn > cut ? ru() : unif_rand();
return floor(dn * u);
}
On Wed, Sep 19, 2018 at 9:20 AM Duncan Murdoch <
murdoch.duncan at gmail.com
<mailto:murdoch.duncan at gmail.com>> wrote:
On 19/09/2018 12:09 PM, Philip B. Stark wrote:
> The 53 bits only encode at most 2^{32} possible values, because
> source of the float is the output of a 32-bit PRNG (the obsolete
> of MT). 53 bits isn't the relevant number here.
No, two calls to unif_rand() are used. There are two 32 bit values,
but
some of the bits are thrown away.
Duncan Murdoch
>
> The selection ratios can get close to 2. Computer scientists
> the way R does, for a reason.
>
> Regards,
> Philip
>
> On Wed, Sep 19, 2018 at 9:05 AM Duncan Murdoch
<murdoch.duncan at gmail.com <mailto:murdoch.duncan at gmail.com>
> <mailto:murdoch.duncan at gmail.com
<mailto:murdoch.duncan at gmail.com>>> wrote:
>
> On 19/09/2018 9:09 AM, I?aki Ucar wrote:
> > El mi?., 19 sept. 2018 a las 14:43, Duncan Murdoch
> > (<murdoch.duncan at gmail.com
<mailto:murdoch.duncan at gmail.com> <mailto:murdoch.duncan at gmail.com
<mailto:murdoch.duncan at gmail.com>>>)
> >>
> >> On 18/09/2018 5:46 PM, Carl Boettiger wrote:
> >>> Dear list,
> >>>
> >>> It looks to me that R samples random integers using an
> >>> algorithm by going from a random number on [0,1) from
> >>> integer, e.g.
> >>>
> >>>
> >>> Many other languages use various rejection sampling
> >>> provide an unbiased method for sampling, such as in Go,
> >>> algorithm currently used in R is also described
> >>> in this area, but does it make sense for the R to adopt
> >>> random sample algorithms outlined there and used in
> >>> a patch providing such an algorithm be welcome? What
> >>> be addressed first?
> >>>
> >>> I believe this issue was also raised by Killie & Philip
> >>> pointing to the python implementation for comparison:
> >>>
> >>
> >> I think the analyses are correct, but I doubt if a change
> >> is likely to be accepted as it would make it more
> >> older results.
> >>
> >> On the other hand, a contribution of a new function like
> >> not suffering from the bias would be good. The normal
> >> a contribution is in a user contributed package.
> >>
> >> By the way, R code illustrating the bias is probably not
> >> put together. I believe the bias manifests itself in
> >> values with two different probabilities (instead of all
> >> probabilities). Those may differ by as much as one part
> >
> > According to Kellie and Philip, in the attachment of the
> > referenced by Carl, "The maximum ratio of selection
> > get as large as 1.5 if n is just below 2^31".
>
> Sorry, I didn't write very well. I meant to say that the
> probabilities would be 2^-32, not that the ratio of
> be 1 + 2^-32.
>
> By the way, I don't see the statement giving the ratio as
> maybe
> I was looking in the wrong place. In Theorem 1 of the paper
> looking in the ratio was "1 + m 2^{-w + 1}". In that formula
> n. If it is near 2^31, R uses w = 57 random bits, so the
> would be
> very, very small (one part in 2^25).
>
> The worst case for R would happen when m is just below
> is at least 31 for the default generators. In that case the
> could
> be about 1.03.
>
> Duncan Murdoch
>
>
>
> --
> Philip B. Stark | Associate Dean, Mathematical and Physical
> Professor, Department of Statistics |
> University of California
> Berkeley, CA 94720-3860 | 510-394-5077 |
--
Philip B. Stark | Associate Dean, Mathematical and Physical Sciences |
Professor, Department of Statistics |
University of California
Berkeley, CA 94720-3860 | 510-394-5077 | statistics.berkeley.edu/~stark
<http://statistics.berkeley.edu/%7Estark> |
@philipbstark
--
Philip B. Stark | Associate Dean, Mathematical and Physical Sciences |
Professor, Department of Statistics |
University of California
Berkeley, CA 94720-3860 | 510-394-5077 | statistics.berkeley.edu/~stark |
@philipbstark