[R] Random seed

Jeff Newmiller jdnewmil @ending from dcn@d@vi@@c@@u@
Mon Dec 24 00:45:31 CET 2018


For the record, low bitcount RNGs are relatively sensitive to initialization [1] such that rules like the one you reference might make sense for specific designs, but the default RNG used in R is much much more sophisticated than that [2].

[1] http://daviddeley.com/random/random4.htm
[2] ?Random (https://stat.ethz.ch/R-manual/R-devel/library/base/html/Random.html)

On December 22, 2018 8:09:42 PM PST, Jeff Newmiller <jdnewmil using dcn.davis.ca.us> wrote:
>It doesn't matter. The whole point is to make the pseudo-random
>sequence repeatable... unless you have a specific reason to avoid
>repeatability.
>
>On December 22, 2018 5:33:39 PM PST, Steven Yen <styen using ntu.edu.tw>
>wrote:
>>I have known from the old days to set a random seed of a LARGE ODD 
>>NUMBER. Now I read instructions of set.seed and it requires ANY
>>INTEGER. 
>>Any idea? Or, does it matter. Thanks.

-- 
Sent from my phone. Please excuse my brevity.



More information about the R-help mailing list