Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. 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

"Let $(\Omega, \mathcal{F}, \mathcal{P})$ be a probability space..." is a typical phrase found in scientific publications. There are a couple of questions regarding this notation.

  1. First, about the probability measure $\mathcal{P}: \Omega \to [0, 1]$. My confusion comes from the fact that, after such an introduction, authors typically start operating on different random variables, which have some particular distribution functions, say $X_i(\omega) \sim \Phi_{X_i}(x)$. $\Phi_{X_i}(x)$, $\forall i$, seem (to me) to be related to different probability measures, which cannot be covered by a signle one, $\mathcal{P}$. Therefore, for each $X_i(\omega)$ I would expect one to define a separate probability space $(\Omega, \mathcal{F}, \mathcal{P}_i)$ such that $F_{X_i}(x) = \mathcal{P}_i(X_i(\omega) \leq x)$. However, I do not see anything like this. There is, probabily, some misunderstanding from my side.

  2. Second, about the corresponding space of square-integrable random variables. Usually, it is denoted by $L^2(\dots)$ with different variations of what is in the brackets. I suppose "square-integrable" makes little sense without a measure; therefore, it should be at least $L^2(\Omega, \mathcal{P})$, but quite often one can find just $L^2(\Omega)$.

Can anyone please comment on this? Thank you.

Regards, Ivan

share|cite|improve this question
up vote 4 down vote accepted
  1. Actually $\mathcal{P}$ is a map $\mathcal{F} \to [0,1]$, not $\Omega \to [0,1]$, and if $X : \Omega \to \mathbb{R}$ is a random variable then $\mathcal{P}(X \le x)$ is actually an 'abuse of notation' used to mean $$\mathcal{P}(X \le x) := \mathcal{P}( \left\{ \omega \in \Omega\, :\, X(\omega) \le x \right\} ) = \mathcal{P} \circ X^{-1}((-\infty, x])$$ and so $\mathcal{P}$ does not depend on the distribution of $X$. [Notice that $(-\infty, x]$ is a Borel-measurable subset of $\mathbb{R}$, so $X^{-1}((-\infty, x])$ is $\mathcal{P}$-measurable (i.e. it lies in $\mathcal{F}$), and so the expression makes sense.]

  2. Square-integrable functions are defined over a probability (or measure) space, so in fact it is $L^2(\Omega, \mathcal{F}, \mathcal{P})$, though sometimes $L^2(\Omega)$ or $L^2(\mathcal{P})$, or even $L^2$, are used as shorthands when the other parameters are understood from context.

share|cite|improve this answer
Thanks a lot, Clive. – Ivan Sep 16 '12 at 20:33

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.