There are many tests of normality that might be well suited. The
Kolmogorov-Smirnov test (
http://en.wikipedia.org/wiki/Kolmogorov%E2%80%93Smirnov_test) for instance
should be easy to compute in terms of the function erf().




On Mon, Dec 10, 2012 at 7:07 PM, Darren Cook <[email protected]> wrote:

> > How much [effort] to determine whether there are multiple peaks?
>
> The Shapiro-Wilk test can give you a probability of how non-normal the
> distribution is:
>   http://en.wikipedia.org/wiki/Shapiro%E2%80%93Wilk_test
>
> As an R example, here is some test data:
>   set.seed(7);
>   data <- c(rnorm(2000,0,40),rnorm(2500, 0, 20), rnorm(400, 40, 5));
>   hist(data,breaks=200)
>
> and running shapiro.test(data) gives me:
>   W = 0.9939, p-value = 1.184e-13
>
> The lower the p-value the more it thinks it is not a normal curve. The
> extreme result is interesting, as the graph looks "roughly normal" to me.
>
> (The Wikipedia page lists alternative tests, which can be found in the R
> nortest package apparently. I've no idea of CPU effort required for each
> of them.)
>
> > Now the tough question: How can this information be used to improve move
> selection?
>
> One approach, not at all sophisticated, is better time management: spend
> less time on normal distributions, more time when the distribution is
> messy.  (But I wonder if more time will just make the two peaks stand
> out more?)
>
> Darren
>
> --
> Darren Cook, Software Researcher/Developer
>
> http://dcook.org/work/ (About me and my work)
> http://dcook.org/blogs.html (My blogs and articles)
> _______________________________________________
> Computer-go mailing list
> [email protected]
> http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
>
_______________________________________________
Computer-go mailing list
[email protected]
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go

Reply via email to