BTW, Ken Kleinman recently wrote a post on how to get a "real" random
numbers (into R) from a web-service:
http://www.r-bloggers.com/example-8-35-grab-true-not-pseudo-random-numbers-passing-api-urls-to-functions-or-macros/
Cheers,
Tal
----------------Contact
Details:-------------------------------------------------------
Contact me: Tal.Galili@gmail.com | 972-52-7275845
Read me: www.talgalili.com (Hebrew) | www.biostatistics.co.il (Hebrew) |
www.r-statistics.com (English)
----------------------------------------------------------------------------------------------
On Fri, Apr 22, 2011 at 6:47 AM, Joshua Wiley wrote:
> On Thu, Apr 21, 2011 at 8:34 PM, Penny Bilton
> wrote:
> > Hi Josh,
> >
> > Thanks for your reply.
> >
> > The problem is have is in trying to retain the proportions of 2 groups in
> my
> > data while sampling into training and test sets. I find that different
> > arguments for set.seed give very different proportions of my 2 groups in
> > the training and test sets.
>
> Sure, just because numbers are random does not guarantee that equal
> numbers from both groups will be sampled. Perhaps you are looking for
> some sort of constrained random sampling like sampling x from group 1
> and x from group 2? If so, try calling sample() separately on each
> group (for help applying the same function to different groups, take a
> look at ?by or ?tapply for example).
>
> Josh
>
> PS cced back to list
>
> >
> >
> > Penny.
> >
> >
> >
> > On 22/04/2011 3:27 p.m., Joshua Wiley wrote:
> >>
> >> Hi,
> >>
> >> On Thu, Apr 21, 2011 at 8:18 PM, Penny Bilton
> >> wrote:
> >>>
> >>> I am using /set.seed()/ before the /sample/ function.
> >>>
> >>> How does the length of the argument of /set.seed()/ and order of the
> >>> digits affect how the sampling is carried out?
> >>
> >> You can use set.seed() to specify a particular seed so that while
> >> pseudo-random numbers are sampled, you can repeat it. For example:
> >>
> >> set.seed(10)
> >> rnorm(10)
> >> set.seed(10)
> >> rnorm(10)
> >>
> >>> Specifically, I have used set.seed(123456789). Will this configuration
> >>> give me a genuinely random sampling??
> >>
> >> You will never get truly random sampling from a computer algorithm,
> >> but it is darn close and more than adequate in the majority of cases.
> >> 123456789 is just a length 1 vector containing the number 123456789,
> >> not 9 separate numbers.
> >>
> >> Google will be able to give you a lot of information on pseudo-random
> >> number algorithms as well as the concept of "seeds". Also see
> >> ?set.seed
> >>
> >> Cheers,
> >>
> >> Josh
> >>
> >>>
> >>> Thank you in anticipation.
> >>>
> >>> Penny.
> >>>
> >>>
> >>> [[alternative HTML version deleted]]
> >>>
> >>> ______________________________________________
> >>> R-help@r-project.org mailing list
> >>> https://stat.ethz.ch/mailman/listinfo/r-help
> >>> PLEASE do read the posting guide
> >>> http://www.R-project.org/posting-guide.html
> >>> and provide commented, minimal, self-contained, reproducible code.
> >>>
> >>
> >>
> >
> >
>
>
>
> --
> Joshua Wiley
> Ph.D. Student, Health Psychology
> University of California, Los Angeles
> http://www.joshuawiley.com/
>
> ______________________________________________
> R-help@r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide
> http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
>
[[alternative HTML version deleted]]