Difference between 32-bit and 64-bit, as was noticed (long) before.
R.E. Boss > -----Oorspronkelijk bericht----- > Van: [email protected] > [mailto:[email protected]] Namens Ric Sherlock > Verzonden: woensdag 12 december 2012 3:05 > Aan: Programming JForum > Onderwerp: Re: [Jprogramming] convert 0,1 coding to _1,1 > > According to the Dictionary ?. uses a fixed random seed ( 7^5 ) and a fixed > RNG so it should always give the same result no matter how many uses of ? > prior to its use. > > I suspect that any differences observed from one machine/session to another > are caused by differences in the bitness (32 vs 64) or differences in the > underlying processor architecture (Intel vs AMD). > > > On Wed, Dec 12, 2012 at 1:56 PM, Linda Alvord <[email protected]>wrote: > > > I think the first digit of ?. locks in a seed and a sequence of random > > digits and then that sequence will be used in all examples in that session > > each time ? is used. > > If 30 random numbers are used in an application, when another application > > is used that session.o > > > > To get "new" random numbers, use ? , but if you now use ?. you can't > > guarantee that someone will have done the same number of uses of ? > > already. > > Thus there is no guarantee that thy will agree. However, you can now use > > that new seed until you change from ?. back to "real random". > > > > I hope this true and really what happens, I'm sure there will be someone > > who > > will set the record straight. > > > > Linda > > > > -----Original Message----- > > From: [email protected] > > [mailto:[email protected]] On Behalf Of Brian > > Schott > > Sent: Tuesday, December 11, 2012 6:28 PM > > To: [email protected] > > Subject: Re: [Jprogramming] convert 0,1 coding to _1,1 > > > > Linda, > > > > Yes, I get your same result on my iPad. > > But, I do not get the same result when instead of 1000, 2 is used on my > > iPad > > and my desktop. That is peculiar, but ... . > > > > On Tue, Dec 11, 2012 at 5:12 PM, Linda Alvord <[email protected]> > > wrote: > > > Try just the first line of this in a new version (as the first use of > > > ?.) in J701jhs or J701gtk and or any old J's you can find and the > > > result should be the same: > > > > > > ]A=.?.3#1000 > > > 146 755 79 > > > > > > > > > > -- > > (B=) > > ---------------------------------------------------------------------- > > For information about J forums see http://www.jsoftware.com/forums.htm > > > > ---------------------------------------------------------------------- > > For information about J forums see http://www.jsoftware.com/forums.htm > > > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
