When generating two separate random numbers, such as for colours for foreground/background or on/off, etc.; it is desirable that the two randomly generated numbers not equal one another. I have come up with this solution, but wonder whether there may be a better/more efficient way of doing it?
This uses an intermediate variable (memory only, not in .ini file), RandColorOn. The generated numbers range, in this case, from 1 to 7 for on/foreground, and from 0 to 7 for off/background.
Oh, I see what you are doing! I had been thinking that if the equal random number was regenerated, it could still come up equal again, but I was forgetting that UniqueRandom=1 would not allow that. So yes, your code is a bit shorter, and no need for an intermediate variable. I like it!
qwerky wrote: ↑March 5th, 2019, 12:58 am
Oh, I see what you are doing! I had been thinking that if the equal random number was regenerated, it could still come up equal again, but I was forgetting that UniqueRandom=1 would not allow that. So yes, your code is a bit shorter, and no need for an intermediate variable. I like it!
Great. Yeah, while statistically the odds of say rolling snake-eyes 10 times in a row is very low, it's not zero. The odds of drawing cards from a deck and discarding them, and drawing the Ace of Spades twice in a row is in fact zero.
It is UniqueRandom that changes this from the kind of random you get with dice or a coin, into the kind of random you get with a deck of cards.
The caveat being that it is possible to draw the Ace of Spades as the last card in that deck, and the first card after you reshuffle and start over. So while nothing in this will entirely prevent either of the numbers being the same twice (and only twice) in a row, the UniqueRandom will make that relatively less likely, and the IfCondition will entirely prevent both of them begin the same.
The effectiveness of UniqueRandom increases as the size of the deck you are drawning from increases... The odds of that "last / first" scenario for a single number happening decreases in an exponential way, although it can never be zero. I don't think there is a practical way to have both conditions be strictly enforced, that a single number can never repeat even once, and both numbers can never be the same. The two "rules" would kinda chase each other around, as "fixing" one could "break" the other.
Ha ha... yes, I thought of that "first/last" (good description) scenario last night--the number that comes up last in the sequence, could come up again as the first one in the new sequence--but then, the IfCondition=Rand2 = Rand1 would just cause the second measure to generate another number, which of course cannot, due to UniqueRandom=1, be the same again, since the sequence is just beginning. Unless LowBound = HighBound, which then no longer gives a random number, but a fixed number, and is pretty meaningless in the context of a random measure!
