cottrell@nbs-vms.ARPA (COTTRELL, JAMES) (10/02/85)
I'll answer the easy part... > 1. What good is a random number generator when you have to generate a > random seed to begin with. For debugging. If the same series is generated each time (by the same seed) then the program will behave in exactly the same way each time it is executed. Remember the `randomize' statement in BASIC? That's what it does, sets the seed to the system clock to look more random. > Are (1), (2), and (3) wrong, am I totally out to lunch... If you are, bring me back a cheeseburger. Hold the randomness. > are there any good references??? Yes, but I'm not one of them. jim cottrell@nbs "Well you wore out your welcome with random precision" -- Pink Floyd */ ------
mikel@codas.UUCP (Mikel Manitius) (10/08/85)
> I'll answer the easy part... > > 1. What good is a random number generator when you have to generate a > > random seed to begin with. > > For debugging. If the same series is generated each time (by the same > seed) then the program will behave in exactly the same way each time > it is executed. Remember the `randomize' statement in BASIC? That's > what it does, sets the seed to the system clock to look more random. Perhaps this bellongs in net.math, or net.philosophy, but since the topic is already being discussed here... I have heard many times that people like the number 17, because it is "the most random number", I find this statement illogical, as how can one number be "more random" than another, unless they mean that this number will occurr most offten when generating "random numbers". Can anyone speculate? -- ======= Mikel Manitius ==----===== AT&T (305) 869-2462 RNX: 755 ==------===== Information Systems ...{akguc|ihnp4}!codas!mikel ===----====== SDSS Regional Support ...attmail!mmanitius =========== Altamonte Springs, FL My opinions are my own. =======
gww@aphasia.UUCP (George Williams) (10/12/85)
> I have heard many times that people like the number 17, because it is > "the most random number", I find this statement illogical, as how can > one number be "more random" than another, unless they mean that this > number will occurr most offten when generating "random numbers". I too have heard 17 called the most random number, but this was always done in a joking fashion, by people who knew they were talking gibberish. I was also given a "proof" that went something like this. Obviously the most random number must be less than 20 (other numbers just don't occur that often). Obviously it most be prime since if it is divisible by two things it can't be random. Well now 2 isn't random because it's even 3 isn't random because it's the smallest odd prime 5 isn't random because it's half of ten 7 I have forgotten 11 isn't random because it's digits repeat 13 I have forgotten 19 is one less than 20 so it isn't random Or something like that (remember :-).
putnam@steinmetz.UUCP (jefu) (10/12/85)
In article <170@codas.UUCP> mikel@codas.UUCP (Mikel Manitius) writes: >I have heard many times that people like the number 17, because it is >"the most random number", I find this statement illogical, as how can >one number be "more random" than another, unless they mean that this >number will occurr most offten when generating "random numbers". 17 is not the most random number, it is the ONLY random number. The proof is by enumeration of possible random numbers... Clearly any number 10 or less is too small. Equally clearly any number 20 or greater is too large. Irrationals are out, as they are, well, irrational. Fractions are hardly random, as the number of possible representations biases things too much. So, we are left with the integers 11-19. 11 is the same digit repeated and this is obviously not good. 12 is the product of two consecutive integers and therefore is special and cant be random. 13 is unlucky, and who would ever want an unlucky random number. 14 is more complicated. 1+4=5, 14=7*2, 7-2=5. And this is far too interesting a property to ever allow 14 to be random. 15 is the product of the first two odd primes. Clearly out. 16 is a square. Nope. 17 is random. 18 is 9*2 and thus has many interesting bad (for randomness) properties. 19 is too close to 20. Thus, 17 is random, and is the only random number. If you remain unconvinced, The three great proof techniques are canonically invoked. First. Proof by dudgeon. "Believe or ill have a screaming fit." Second. Proot by luncheon. "If i buy you lunch will you believe?" Third. Proof by bludgeon. "Believe or i will hit you over the head with this 2 by 4." Now we can begin to generate all the pseudo-random numbers : 34=2*17, 51... .1717171717171717171717... and so forth. Sigh. This didnt (indeed) belong in net.unix-wizards. But i couldnt resist. Followup to net.math. -- O -- jefu tell me all about -- UUCP: {rochester,edison}!steinmetz!putnam Anna Livia! I want to hear all.... -- ARPA: putnam@GE-CRD
jqj@cornell.UUCP (J Q Johnson) (10/13/85)
> 17 is the most random number.
Similarly, of course, it is well known that there are no uninteresting
natural numbers, for if there were then there would be a first one, and
the property "the first uninteresting natural number" is certainly
interesting. Long live the real numbers! Down with floating point
numbers, which are just a finite subset of the naturals in disguise!