Skip to content

RNG initialization

3 messages · Paul Gilbert, Brian Ripley

Original
Brian Ripley
# 
and
Sure, it is easy and that was what I was proposing to do. I was just
proposing to be compatible with S-PLUS and not syskern.  I really don't
see why syskern needs to use the name set.seed.
Paul Gilbert
# 
No, syskern does not need to use the name set.seed. I used it hoping this stuff
would move to R  base at some point and I could remove it from syskern.

Beyond the name however, there are some Splus compatiblity issues which probably
should be discussed. I seem to recall that there was some difficulty producing
Splus runif results in R.  Also, the ability to specify the RNG in R is nice and
should be accommodated, and setting the seed without setting the RNG is probably
not wise. Finally, I feel it is important to also specify the algorithms for
other transformations such rnorm. This way, one can be relatively confident that
a random experiment can be reproduced.

Paul Gilbert

-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-devel mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To: r-devel-request@stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._
Brian Ripley
#
On Tue, 18 Jan 2000, Paul Gilbert wrote:

            

      
      
      
    

        
Of course, as the Splus runif is not implemented in R. There is something
similar, but it is not the same.