I've only formally studied probability math to the point of combinations and permutations, so some of the statistics concepts and terms get confusing to me. Perhaps I'm asking an obvious question. If I were to have a constant probability at every step, giving me a geometric distribution, will...
if the first curve would represent the probability of stopping on every step, and the second would represent the number of trials that stopped on every step.
The most simple form of randomness I have is the built-in random function which gives me a pseudo-random number in a specified range. I was assuming that a constant probability would give me a uniform distribution where the process will stop at the first iterations equally often as it will run...
Yes, I think that is correct if I understand it right.
As for the mean and the standard deviation, I adjusted the standard normal distribution CDF values to fit my normal distribution using (duration - 20)/20 * 3.09), perhaps that messed it up?
I'm trying to code a sub-routine for a simple...
Thing is, I think I already know a different way to achieve the desired results by calculating the runtime for every process before it starts.
I would just have the computer generate random values in 0.0-1.0 and apply the Box-Muller to them, or some better transform, giving me a...
Well, all I really want is for the distribution of the different runtimes to have a bell-curve shape like in the picture on the right here: http://en.wikipedia.org/wiki/Normal_distribution
Whichever implementation is the simplest and gives me the most control.
my CDF function currently just...
I'm programming a process that runs in iterations with a probability of stopping at every iteration. As it gets closer to the desired run-time, this probability should increase so that if the process is run many times over, the runtimes will form a normal distribution, with most of them having...