Definition of liminf of sequence of functions?

mathmonkey
Messages
33
Reaction score
0

Homework Statement



Hi I've come across the term lim inf ##f_n## in my text but am not sure what it means.

##\lim \inf f_n = \sup _n \inf _{k \geq n} f_k##

In fact, I am not sure what is supposed to be the output of lim inf f? That is, is it supposed to return a real-valued number, or a function itself?

Generally for a real-valued function ##\inf f## refers to ##\inf_x f(x)##. That is, it returns the largest real-valued number smaller than f(x) for all x. If that's the case, then it should follow that ##\lim \inf f_n## also returns a real-valued number? But I've always thought the implication of lim inf and lim sup is that if ##f_n## converges uniformly to ##f## then

##\lim \inf f_n = \lim \sup f_n = \lim f_n = f##

but that doesn't seem to hold if i use this definition? Any help or clarification would be greatly appreciated. Thanks!
 
Physics news on Phys.org
mathmonkey said:

Homework Statement



Hi I've come across the term lim inf ##f_n## in my text but am not sure what it means.
I really like this image, from Wikipedia. It helped get me started in understanding intuitively what limsup and liminf mean. Hopefully it does the same for you.
 
Hi Mandelbroth,

Thanks for your reply! However, that picture only describes the limsup/liminf of a sequence of points, which is itself a point, which is easier to intuit for me. But what I'm wondering is what is the limsup/liminf of a sequence of functions supposed to be? Is it to be a function itself? Or just a point? I am unclear as to the definition itself of limsup for sequences of functions...

So far I'm thinking it shouldn't be intuitive for limsup/liminf f_n to be a point, since as I mentioned earlier if ##f_n## converges to ##f## uniformly it ought to be the case that ##\lim \sup f_n = \lim \inf f_n = \lim f_n##? Thanks again for any help
 
mathmonkey said:
Hi Mandelbroth,

Thanks for your reply! However, that picture only describes the limsup/liminf of a sequence of points, which is itself a point, which is easier to intuit for me. But what I'm wondering is what is the limsup/liminf of a sequence of functions supposed to be? Is it to be a function itself? Or just a point? I am unclear as to the definition itself of limsup for sequences of functions...

So far I'm thinking it shouldn't be intuitive for limsup/liminf f_n to be a point, since as I mentioned earlier if ##f_n## converges to ##f## uniformly it ought to be the case that ##\lim \sup f_n = \lim \inf f_n = \lim f_n##? Thanks again for any help

Yes, it's a function. The value of lim inf ##f_n## at a point x is the lim inf of sequence of numbers ##f_n(x)## as n->infinity.
 
mathmonkey said:
Hi Mandelbroth,

Thanks for your reply! However, that picture only describes the limsup/liminf of a sequence of points, which is itself a point, which is easier to intuit for me. But what I'm wondering is what is the limsup/liminf of a sequence of functions supposed to be? Is it to be a function itself? Or just a point? I am unclear as to the definition itself of limsup for sequences of functions...

If you understand limsup/liminf of sequences of points, then this isn't too hard. Basically, you take a sequence of functions ##(f_n)_n##. Now, if I take a fixed ##x##, then ##x_n = f_n(x)## is a sequence of points. So the liminf makes sense. Now, we define

f(x) = \liminf x_n

And we do that for any point. So the liminf of a sequence of functions is again a function ##f## which satisfies that

f(x) = \liminf f_n(x)

for any ##x##. So you evaluate the liminf pointswize.

So far I'm thinking it shouldn't be intuitive for limsup/liminf f_n to be a point, since as I mentioned earlier if ##f_n## converges to ##f## uniformly it ought to be the case that ##\lim \sup f_n = \lim \inf f_n = \lim f_n##? Thanks again for any help

This is true. But uniform convergence isn't even needed. Pointswize convergence is enough.
 
Thanks guys! That makes perfect sense!
 
Prove $$\int\limits_0^{\sqrt2/4}\frac{1}{\sqrt{x-x^2}}\arcsin\sqrt{\frac{(x-1)\left(x-1+x\sqrt{9-16x}\right)}{1-2x}} \, \mathrm dx = \frac{\pi^2}{8}.$$ Let $$I = \int\limits_0^{\sqrt 2 / 4}\frac{1}{\sqrt{x-x^2}}\arcsin\sqrt{\frac{(x-1)\left(x-1+x\sqrt{9-16x}\right)}{1-2x}} \, \mathrm dx. \tag{1}$$ The representation integral of ##\arcsin## is $$\arcsin u = \int\limits_{0}^{1} \frac{\mathrm dt}{\sqrt{1-t^2}}, \qquad 0 \leqslant u \leqslant 1.$$ Plugging identity above into ##(1)## with ##u...
Back
Top