I'm trying to apportion a value between two buckets / classify N outcomes such that:

  • X% are "Low Priority"
  • 100% -X % are "High Priority"

I've got a percentage value of X% and a deviation, but I'm struggling with the distribution aspect. Frequently for example, I end up calculating an X that is effectively above 100% when using a normal distribution, a value of X of ~0.95 and a deviation of 0.026. For example, occasionally I'll end up with 102.83% X and thus -2.83% for X', which is clearly not possible in my data.

Logically is a normal distribution the right construct to use here? My other guess was basically to sample up to N times, and for each N randomly select is it low/high based on an absolute scalar value - but this feels like a brutal approach that would perform poorly for large N's?

Questions mainly are:

  • Am I on the right track at all?
  • Should I be using another distribution here in lieu of the normal, to calculate the split/spread of X vs 1-X total?
  • If I'm using an appropriate distribution, how do I correctly handle the wrap-around effect of over-running my range?
  • $\begingroup$ Why are you using a distribution at all? You have some cases, you call (say) 5% of them high-priority — but that does not require a distribution. $\endgroup$
    – Matt F.
    Nov 24 '21 at 2:17

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.