- #1
- 4,807
- 32
In the Riemann theory for a function f defined on all of R, we define its improper integral over R as the sum of two limits:
[tex]\int_{-\infty}^{+\infty}f(x)dx = \lim_{a\rightarrow -\infty}\int_{a}^0f(x)dx+\lim_{b\rightarrow +\infty}\int_{0}^b f(x)dx[/tex]
and in general, this is not equal to
[tex]\lim_{L\rightarrow +\infty}\int_{-L}^L f(x)dx[/tex]
For instance, for f(x)=sin(x), neither of the two limits above exists, but the limit below is 0.
But what does Lebesgue says? We know that given f:R-->R integrable, the function [tex]\nu:\mathfrak{L}_{\mathbb{R}}\rightarrow \mathbb{R}[/tex] defined by
[tex]\nu(E)=\int_Ef(x)dx[/tex]
is a (signed) measure. Signed measures enjoy the same continuity properties as finite and positive measures. That is, if E_n is an increasing sequence of measurable sets, then
[tex]\nu\left(\bigcup_{n=1}^{+\infty}E_n\right)=\lim_{n\rightarrow +\infty}\nu(E_n)[/tex]
In particular, for E_n = ]-n,n[ and the measure defined above,
[tex]\int_{-\infty}^{+\infty}f(x)dx=\lim_{n\rightarrow +\infty}\int_{-n}^nf(x)dx[/tex]
(Since U E_n=R)
What's up with that? (as in : any comments welcome)
[tex]\int_{-\infty}^{+\infty}f(x)dx = \lim_{a\rightarrow -\infty}\int_{a}^0f(x)dx+\lim_{b\rightarrow +\infty}\int_{0}^b f(x)dx[/tex]
and in general, this is not equal to
[tex]\lim_{L\rightarrow +\infty}\int_{-L}^L f(x)dx[/tex]
For instance, for f(x)=sin(x), neither of the two limits above exists, but the limit below is 0.
But what does Lebesgue says? We know that given f:R-->R integrable, the function [tex]\nu:\mathfrak{L}_{\mathbb{R}}\rightarrow \mathbb{R}[/tex] defined by
[tex]\nu(E)=\int_Ef(x)dx[/tex]
is a (signed) measure. Signed measures enjoy the same continuity properties as finite and positive measures. That is, if E_n is an increasing sequence of measurable sets, then
[tex]\nu\left(\bigcup_{n=1}^{+\infty}E_n\right)=\lim_{n\rightarrow +\infty}\nu(E_n)[/tex]
In particular, for E_n = ]-n,n[ and the measure defined above,
[tex]\int_{-\infty}^{+\infty}f(x)dx=\lim_{n\rightarrow +\infty}\int_{-n}^nf(x)dx[/tex]
(Since U E_n=R)
What's up with that? (as in : any comments welcome)
Last edited: