Last active
August 22, 2017 17:25
-
-
Save falkreon/28bf8bed26a9d766b100c1b0cfa06c09 to your computer and use it in GitHub Desktop.
Selecting unusually-distributed pseudorandom numbers
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| /** | |
| * Creates a random number from (-1..1), exclusive, distributed in an inverse-bell-curve fashion. That is, numbers | |
| * closer to -1 or 1 are exponentially more likely to appear than numbers closer to 0. | |
| */ | |
| public static double invertedNormalRandom(Random r) { | |
| /* | |
| * Implementation note: log10 reaches y=0 at x=1, and reaches y=1 at x=10, so it's really important, if we | |
| * want to get good numbers out of it, to feed it numbers in the range of 1..10. So we multiply by 9 and add 1. | |
| */ | |
| double a = Math.log10((r.nextDouble()*9)+1); | |
| return (r.nextBoolean()) ? a : -a; | |
| } |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Also, if you're wondering why I don't just use sine, it creates a thoroughly unpleasant distribution, squished hard to the edges and with a spike in the middle due to the high slope near the center. Sometimes the transformations needed to create a given distribution are extremely unintuitive; even gaussianRandom itself seems to have an anomalous spike at zero which does not fit its uniform distribution claim(!)