MathOverflow is a question and answer site for professional mathematicians. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

I'm looking for a list of examples of random variables to use in teaching a measure-theoretic probability course. For example, the Rademacher functions are an explicit construction of independent Bernoulli random variables.

If you were teaching such a course and had a list of canonical examples to illustrate definitions and theorems, what would be on the list?

share|cite|improve this question
up vote 4 down vote accepted

That will be not quite an answer for your question. Anyway it may be helpful.

  1. If you have a sequence of independent Bernoulli r.v. $(B_i)$ then you can define a uniform variable by $U = \sum 2^{-i} B_i$ and further you can obtain an infinite sequence of independent uniform r.v $U_i$. (just by splitting $B_i$ into infinitely many subsequences). Finally from this sequence you may get a sequence of independet variables of any continuous distribution $F^{-1}(U_i)$ where $F^{-1}$ is the generalised inverse of the cdf of the distribution we want to have.

  2. There is a construction of the Wiener process using Haar functions. I guess it is may be to difficult for your students. But I can look for references in English (in a moment I have only a Polish book).

share|cite|improve this answer

You seem to be asking for examples of random variables realized concretely as measurable functions on a probability space. This runs rather counter to the usual point of view of probability theory, which only cares about the distributions (more generally, joint distributions) of random variables. (In particular, as Piotr points out, it's relatively elementary to construct a sequence of independent random variables with arbitrary distributions, defined on [0,1].)

That comment notwithstanding, here are some examples:

  1. Put the uniform probability measure on {0,1}^n, or more generally the n-fold product of a measure which puts mass p on 1 and mass (1-p) on 0. The function $f(x) = \sum x_i$ is a binomial random variable.

  2. Put the rotation-invariant probability measure on the sphere $\sqrt{n}S^{n-1}$. The function $f(x) = x_1$ is a random variable which converges in distribution to the standard Gaussian distribution when $n\to \infty$. (This is sometimes called the "Poincare limit" because it was first observed by Maxwell and first rigorously proved by Borel.)

  3. Put the uniform probability measure on the permutation group S_n. The number of fixed points of a permutation is a random variable which converges in distribution to a Poisson distribution with mean 1. (Showing that the mean of this random variable is 1 for every n is a nice exercise in the method of indicators.)

share|cite|improve this answer
Example 2 is realy cute. I would like add that if one takes f(x) = (x_1, x_2, ..., x_d) then in the limit one get the d-dimentional random vector of i.i.d. starndard Gaussian variables. – Piotr Miłoś Nov 18 '09 at 12:38

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

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