0
$\begingroup$

Background

I know this question might be a bit weird. More out of curosity.

I know from the definition that one of the criteria for $V$ to be a vector space is that $\forall v_1,v_2 \in V, v_1 + v_2 \in V$.

But how about countable infinite sum? From my viewpoint, it sounds more like a vector space when infinite countable sum is still contained in $V$.

Example

I have this question because when I look at the $l^{\infty}$, i.e., the bounded sequence space. I feel like if I can recursively sum up two proper sequence in $l^{\infty}$, I can make resulting sequence larger than any given bound. So it sounds like, it might be out of the vector space that I am interested in.

I know the definition of vector space requires "two element sum still contained in the space". Just curious if there is a reason for not allowing infinite sum.

$\endgroup$
3
  • 2
    $\begingroup$ To incorporate infinities you need some way of adding meaning to an infinite sum. This leads to a norm, distance or topology. There are plenty of interesting vector spaces without infinite sums ($\mathbb{R}^n$, for example). $\endgroup$
    – copper.hat
    Sep 22, 2018 at 2:05
  • 1
    $\begingroup$ The issue is that we need some form of topology to put on the space to talk about infinite sums. We need to be able to define what convergence means. See both this MSE post asking the same question as you and the wikipedia pages on Metric Spaces, Hilbert Spaces, etc $\endgroup$
    – Brevan Ellefsen
    Sep 22, 2018 at 2:06
  • $\begingroup$ @BrevanEllefsen, this is a great link! $\endgroup$
    – ArtificiallyIntelligence
    Sep 22, 2018 at 2:08

2 Answers 2

Reset to default
6
$\begingroup$

Consider the simple vector space $\mathbb Q$ (rational numbers). It is indeed the case that $$\forall q_1, q_2 \in \mathbb Q : q_1+q_2 \in \mathbb Q.$$

However,

$$1 + \frac{1}{2} + \frac{1}{6} + \frac{1}{24} + \frac{1}{120} + \cdots + \frac{1}{n!} + \cdots = e \notin \mathbb Q.$$

So, infinite sums are effectively quite a different beast than finite sums as far as vector space closure is concered.

$\endgroup$
2
$\begingroup$

Think of $\Bbb R$, which we all know and love. It is a fine vector space. We know how to add any two reals, which tells us how to add any finite collection of reals. We can't use that to define the sum of an infinite sequence of reals. We go through all sorts of machinations to define the sum of some infinite sequences of reals, which we call convergent. The rest we can't give meaning to. What would you like the sum of $1+1+1+1+\ldots $ to be?

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.