Would http://rosettacode.org/wiki/Linear_congruential_generator#J satisfy your need for a "portable" "RNG" or do you have deeper needs?
Thanks, -- Raul On Fri, Jan 24, 2014 at 9:28 AM, Brian Schott <[email protected]> wrote: > If ? had the rank of ?. , it would still behave as ? , if these examples > are relevant. > > ?."? 2 10 $100 > 94 94 94 94 94 94 94 94 94 94 > 94 94 94 94 94 94 94 94 94 94 > ?"?. 2 10 $100 > 54 83 5 96 96 31 83 59 79 54 > 81 80 49 68 32 12 92 66 50 92 > > But on a similar, but different, issue: I have long sought in J a > *portable* RNG (although my research has veered away from this pursuit). By > portable, I mean one that can produce the same stream of numbers on any > modern computer, regardless of the word length. For an extreme example, on > my iPad, the first example produces all 46's, not 94's. The links below > discuss this issue. > > http://www.jstor.org/discover/10.2307/2347988 > http://degiorgi.math.hr/aaa_sem/Rand_Gen/p132-schrage.pdf > > > > On Fri, Jan 24, 2014 at 8:38 AM, Ben Gorte - CITG > <[email protected]>wrote: > >> I see... >> >> In fact your question is why monadic ? has rank 0, instead of _ , which >> would make it like behave more like ?. concerning b. 0 . >> >> We agree that a rank 0 ?. wouldn't be useful: >> >> querydot =: (?.)"0 >> querydot (10#100) >> 46 46 46 46 46 46 46 46 46 46 >> querydot"1 (10#100) >> 46 46 46 46 46 46 46 46 46 46 >> >> Is your question (if it is) not the same sort of question as why +, -, * >> etc. have rank 0? >> Or: what's wrong with: >> >> 11 12 13 +"_ i.3 4 >> 11 12 13 14 >> 16 17 18 19 >> 21 22 23 24 >> >> (I don't kow) >> >> Ben >> > -- > (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
