$P({X^{-1}(x)}) = p$ for all $x$ then Field is finite

48 Views Asked by At

Notation:

  • $\Omega$ denotes a $\sigma-$Field

  • $X$ is a function onto $\Omega$

  • $X^{-1}$ is pre-image

  • $P(\cdot)$ is a probability measure

  • $X^{-1}$ is also a $\sigma-$Field, as a consequence of properties of preimage

Claim:

$P(\{ X^{-1}(x)\})=p$ $\;\forall x \in \Omega$
Then $\Omega$ has finite number of points

Logic:

By Definition of PDF: $k \cdot \sum P(\{ X^{-1}(x)\})=1$

Thus, this sum should be definite, as in $k$ should be finite, hence the preimage should be finite ,ie, $\Omega$ should have finite number of points.

Is This reasoning correct?

Edit:

I don't think the original question says anything about probability distributions. So considering $P$ as only a probability measure:
$\cup A_i = \Omega$ for some $A_i$

By Kolmogorov additivity axiom: $P(X^{-1}( \cup A_i)) = \sum P(X^{-1}(A_i))$

$= P(X^{-1}(\Omega)) = 1$

Hence number of elements of $\Omega$ is $n \in \mathbb{N}$ such that $n\cdot p=1$