16
$\begingroup$

What is the meaning of the tilde when specifying probability distributions? For example:

$$Z \sim \mbox{Normal}(0,1).$$

$\endgroup$
  • 7
    $\begingroup$ Have a look at point 4 of this entry from Wolfram MathWorld. $\endgroup$ – user10525 Oct 27 '12 at 15:40
  • 2
    $\begingroup$ @Procrastinator: you should go ahead and submit this as an answer. I don't think it will get any better. $\endgroup$ – Stephan Kolassa Oct 27 '12 at 19:23
15
$\begingroup$

The ~ (tilde) used in that way means "is distributed as". Why? To ask why doesn't make much sense to me, its just a convention. To cite Brian Ripley:

Mathematical conventions are just that, conventions. They differ by field of mathematics. Don't ask us why matrix rows are numbered down but graphs are numbered up the y axis, nor why x comes before y but row before column. But the matrix layout has always seemed illogical to me. -- Brian D. Ripley (answering a question why print(x) and image(x) are layouted differently) R-help (August 2004)

$\endgroup$
  • 1
    $\begingroup$ I'll wait and see if someone comes up with an idea about the history or the "why" and if not i'll accept this one $\endgroup$ – jsj Oct 28 '12 at 2:00
6
$\begingroup$

I can't comment on the history, but I believe it might be the following. The ~ symbol is commonly used in mathematics to denote an equivalence relation. In the context of probability theory it is used to denote equivalance in (marginal) distribution. So when we say,

Z ~ N(0,1),

what we mean is that the random variable Z has the same marginal distribution as the random variable N(0,1). (The latter being a standard normal random variable, by definition.) This interpretation requires that you interpret the right-hand-side of the equation as referring to a random variable, not a distribution function. Under this interpretation, the ~ sign means "has the same distribution as". Since this is reflexive, symmetric and transitive, it is an equivalence relation.

$\endgroup$
  • $\begingroup$ Equivalence relation on what set? There is no such thing as a "set of all random variables." $\endgroup$ – whuber Jan 1 '13 at 4:07
  • $\begingroup$ One probably could have "something like" equivalence relations in the contect of categories, that is, on proper classes. $\endgroup$ – kjetil b halvorsen Apr 5 '17 at 15:18

Your Answer

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

Not the answer you're looking for? Browse other questions tagged or ask your own question.