Lebesgue Measure Invariant under Translations

From ProofWiki
Jump to: navigation, search

Contents

Theorem

Let $\lambda^n$ be the $n$-dimensional Lebesgue measure on $\R^n$ equipped with the Borel $\sigma$-algebra $\mathcal B \left({\R^n}\right)$.

Let $\mathbf x \in \R^n$.


Then $\lambda^n$ is translation-invariant; i.e., for all $B \in \mathcal B \left({\R^n}\right)$, have:

$\lambda^n \left({\mathbf x + B}\right) = \lambda^n \left({B}\right)$

where $\mathbf x + B$ is the set $\left\{{\mathbf x + \mathbf b: \mathbf b \in B}\right\}$.


Proof

Denote with $\tau_{\mathbf x}: \R^n \to \R^n$ the translation by $\mathbf x$.

From Translation in Euclidean Space is Measurable Mapping, $\tau_{\mathbf x}$ is $\mathcal B \left({\R^n}\right) \, / \, \mathcal B \left({\R^n}\right)$-measurable.

Consider the pushforward measure $\lambda^n_{\mathbf x} := \left({\tau_{\mathbf x}}\right)_* \lambda^n$ on $\mathcal B \left({\R^n}\right)$.


By Characterization of Euclidean Borel Sigma-Algebra, it follows that:

$\mathcal B \left({\R^n}\right) = \sigma \left({\mathcal{J}^n_{ho}}\right)$

where $\sigma$ denotes generated $\sigma$-algebra, and $\mathcal{J}^n_{ho}$ is the set of half-open $n$-rectangles.


Let us verify the four conditions for Uniqueness of Measures, applied to $\lambda^n$ and $\lambda^n_{\mathbf x}$.

Condition $(1)$ follows from Half-Open Rectangles Closed under Intersection.

Condition $(2)$ is achieved by the sequence of half-open $n$-rectangles given by:

$J_k := \left[{-k \,.\,.\, k}\right)^n$


For condition $(3)$, let $\left[[{\mathbf a \,.\,.\, \mathbf b}\right)) \in \mathcal{J}^n_{ho}$ be a half-open $n$-rectangle.

Since:

$\tau_{\mathbf x}^{-1} \left({\left[[{\mathbf a \,.\,.\, \mathbf b}\right))}\right) = \mathbf x + \left[[{\mathbf a \,.\,.\, \mathbf b}\right)) = \left[[{\mathbf {a + x} \,.\,.\, \mathbf {b + x}}\right))$

we have:

\(\displaystyle \) \(\displaystyle \) \(\displaystyle \) \(\displaystyle \lambda^n_{\mathbf x} \left({\left[\left[{\mathbf a \,.\,.\, \mathbf b}\right)\right)}\right)\) \(=\) \(\displaystyle \lambda^n \left({\left[\left[{\mathbf {a + x} \,.\,.\, \mathbf {b + x} }\right)\right)}\right)\) \(\displaystyle \) \(\displaystyle \) \(\displaystyle \)          Definition of pushforward measure          
\(\displaystyle \) \(\displaystyle \) \(\displaystyle \) \(\displaystyle \) \(=\) \(\displaystyle \prod_{i \mathop = 1}^n \left({\left({b_i + x_i}\right) - \left({a_i + x_i}\right)}\right)\) \(\displaystyle \) \(\displaystyle \) \(\displaystyle \)          Definition of Lebesgue measure          
\(\displaystyle \) \(\displaystyle \) \(\displaystyle \) \(\displaystyle \) \(=\) \(\displaystyle \prod_{i \mathop = 1}^n \left({b_i - a_i}\right)\) \(\displaystyle \) \(\displaystyle \) \(\displaystyle \)                    
\(\displaystyle \) \(\displaystyle \) \(\displaystyle \) \(\displaystyle \) \(=\) \(\displaystyle \lambda^n \left({\left[\left[{\mathbf a \,.\,.\, \mathbf b}\right)\right)}\right)\) \(\displaystyle \) \(\displaystyle \) \(\displaystyle \)          Definition of Lebesgue measure          

Finally, since:

$\displaystyle \lambda^n \left({J_k}\right) = \prod_{i \mathop = 1}^n \left({k - \left({-k}\right)}\right) = \left({2 k}\right)^n$

the last condition, $(4)$, is also satisfied.


Whence Uniqueness of Measures implies that:

$\lambda^n_{\mathbf x} = \lambda^n$

and since for all $B \in \mathcal B \left({\R^n}\right)$ we have:

$\mathbf x + B = \tau_{\mathbf x}^{-1} \left({B}\right)$

this precisely boils down to:

$\lambda^n \left({\mathbf x + B}\right) = \lambda^n \left({B}\right)$

$\blacksquare$


Note

This theorem formalizes the physical intuition that the size of an object does not depend on its position.


Sources

Personal tools
Namespaces
Variants
Actions
Navigation
ProofWiki.org
ToDo
Toolbox
Google AdSense