I have a group of objects as observations (e.g. people). Each object has its own value of property A (e.g. height).

Without knowing the distribution of property A among these objects, I want to select out a subgroup of objects to form a certain distribution (e.g. a Gaussian). How should I proceed?

Let's imagine the observed distribution is a uniform one, but with occasional missing samples at certain values (e.g. at 180 cm, 175 cm, etc.). It's fine to have missing bins at 175 and 180 cm in the sampled-out-Gaussian-distributed objects. Those can be regarded as statistical insufficiency.


  • $\begingroup$ Without a ton of thought, I believe the hypergeometric distribution is related to this. $\endgroup$ – The Count Jan 3 '17 at 21:18
  • $\begingroup$ maybe divide your sample into smaller smaples and take averages? $\endgroup$ – Markoff Chainz Jan 3 '17 at 21:24
  • $\begingroup$ Markoff's idea doesn't match to the situation I'm afraid $\endgroup$ – frankliuao Jan 4 '17 at 17:56
  • $\begingroup$ The Count, I don't know how that applies $\endgroup$ – frankliuao Jan 4 '17 at 17:57

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.