Register to reply

Proof analysis

by evagelos
Tags: analysis, proof
Share this thread:
evagelos
#1
Jun6-11, 07:42 AM
P: 308
Suppose we are asked to prove :

If [tex]\lim_{n\to\infty}x_{n} = L[/tex] and [tex]x_{n}[/tex] is decreasing,then [tex]\forall n[n\in N\Longrightarrow x_{n}\geq L][/tex].

On the following proof give a list of all the thoerems ,axioms,definitions and the laws of logic that take place in the proof


proof:
Suppose that there exists [tex]n_1\in\mathbb N[/tex] such that [tex]x_{n_1}<L[/tex]. Take [tex]\epsilon=L-x_{n_1}[/tex]. Since [tex]\displaystyle\lim_{n\rightarrow\infty}x_n=L[/tex] there exists [tex]n_2\in\mathbb N[/tex] such that [tex]\|x_n-L\|<\epsilon[/tex] for all [tex]n\geq n_2[/tex]. Take [tex]n_0>\max\{n_1,n_2\}[/tex]. Then we have that [tex]x_{n_1}<x_n<2L-x_{n_1}[/tex] for all [tex]n\geq n_0[/tex] in particular, [tex]x_{n_1}<x_{n_0}[/tex]. But since [tex]n_1<n_0[/tex] and the sequence is decreasing we have that [tex]x_{n_0}<x_{n_1}[/tex] wich gives us a contradiction. Hence [tex]x_n\geq L[/tex] for all [tex]n\in\mathbb N[/tex].
Phys.Org News Partner Science news on Phys.org
Suddenly, the sun is eerily quiet: Where did the sunspots go?
'Moral victories' might spare you from losing again
Mammoth and mastodon behavior was less roam, more stay at home
micromass
#2
Jun6-11, 08:18 AM
Mentor
micromass's Avatar
P: 18,017
Hi evagelos!

Let's start with

Quote Quote by evagelos View Post
proof:
Suppose that there exists [tex]n_1\in\mathbb N[/tex] such that [tex]x_{n_1}<L[/tex]. Take [tex]\epsilon=L-x_{n_1}[/tex].
What things do you need for this?
evagelos
#3
Jun6-11, 09:25 AM
P: 308
Quote Quote by micromass View Post
Hi evagelos!

Let's start with



What things do you need for this?
No axioms no theorems ,only the definition definition:

[tex]\forall\epsilon[ \epsilon>0\Longrightarrow\exists k(k\in N\wedge\forall n(n\geq k\Longrightarrow |x_{n}-L|<\epsilon))][/tex].

The question now is,what laws of logic

micromass
#4
Jun6-11, 09:47 AM
Mentor
micromass's Avatar
P: 18,017
Proof analysis

Quote Quote by evagelos View Post
No axioms no theorems ,only the definition definition:

[tex]\forall\epsilon[ \epsilon>0\Longrightarrow\exists k(k\in N\wedge\forall n(n\geq k\Longrightarrow |x_{n}-L|<\epsilon))][/tex].
We didn't use convergence yet. So that's not the definition we need.
We do use some definitions and axioms here. For example, you use [itex]\mathbb{N}[/itex], you need to define this in some way and there are quite a lot of axioms that you need to define the natural numbers.

Furthermore, you use the inequality relation < and the difference operation -. In order for these to be well-defined, you had to use some axioms and definitions.

And what about [itex]x_{n_1}[/itex] this is an element of a sequence, that is you have a function [itex]x:\mathbb{N}\rightarrow \mathbb{R}[/itex], but what is a function? You'll need axioms and definitions to precisely define functions. Furthermore, in this function, the codomain is [itex]\mathbb{R}[/itex], and this set needs some axioms and definitions too.

The question now is,what laws of logic
Since we did not do any logical derivations yet, we didn't use much logical laws yet!
evagelos
#5
Jun6-11, 10:23 AM
P: 308
Quote Quote by micromass View Post
We didn't use convergence yet. So that's not the definition we need.
We do use some definitions and axioms here. For example, you use [itex]\mathbb{Nitex], you need to define this in some way and there are quite a lot of axioms that you need to define the natural numbers.

Furthermore, you use the inequality relation < and the difference operation -. In order for these to be well-defined, you had to use some axioms and definitions.

And what about [itex]x_{n_1}[/itex] this is an element of a sequence, that is you have a function [itex]x:\mathbb{N}\rightarrow \mathbb{R}[/itex], but what is a function? You'll need axioms and definitions to precisely define functions. Furthermore, in this function, the codomain is [itex]\mathbb{R}[/itex], and this set needs some axioms and definitions too.



Since we did not do any logical derivations yet, we didn't use much logical laws yet!
O,k then you state the axioms and definitions involved here ,because i am lost.

But surely the application of the laws of logic uppon those axioms and definitions should give us the statement of the proof
micromass
#6
Jun6-11, 10:59 AM
Mentor
micromass's Avatar
P: 18,017
Well, how did we define the natural numbers? And what axioms do you need for that?
An excellent reference for this is staff.science.uva.nl/~vervoort/AST/ast.ps but you'll need to be able to open .ps files...
evagelos
#7
Jun6-11, 11:26 AM
P: 308
Quote Quote by micromass View Post
Well, how did we define the natural numbers? And what axioms do you need for that?
An excellent reference for this is staff.science.uva.nl/~vervoort/AST/ast.ps but you'll need to be able to open .ps files...
Micromass like that we are getting nowhere.

You mean every time we mention the Natural Nos we have to write down their axioms ??

What for ?
micromass
#8
Jun6-11, 11:38 AM
Mentor
micromass's Avatar
P: 18,017
Quote Quote by evagelos View Post
Micromass like that we are getting nowhere.

You mean every time we mention the Natural Nos we have to write down their axioms ??

What for ?
Judging from your post history, you like it that way. If you don't want it that formal, then you'll have to post from what axioms and definitions you're starting. If you want to take the natural numbers for granted, fine. But then you'll have to say what you're taking for granted and what not...
evagelos
#9
Jun7-11, 10:08 AM
P: 308
Quote Quote by micromass View Post
Judging from your post history, you like it that way. If you don't want it that formal, then you'll have to post from what axioms and definitions you're starting. If you want to take the natural numbers for granted, fine. But then you'll have to say what you're taking for granted and what not...

O.k

Do you agree that :[tex]\exists n_{1}[n_{1}\in N\wedge x_{n_{1}}< L][/tex] is the negation of the statement:

[tex]\forall n[n\in N\Longrightarrow x_{n}\geq L][/tex] ??
micromass
#10
Jun7-11, 10:47 AM
Mentor
micromass's Avatar
P: 18,017
Yes, of course.
evagelos
#11
Jun7-11, 02:59 PM
P: 308
Quote Quote by micromass View Post
Yes, of course.
Hence so far we have no axioms or thoerem or definitions involved.

Do you agree??
micromass
#12
Jun7-11, 03:02 PM
Mentor
micromass's Avatar
P: 18,017
Not exactly! The converse of [itex]x_n< L[/itex] is not always [itex]x_n\geq L[/itex]. We have used here that the order relation is total. I.e. that for every a and b, we have

[tex]a\leq b~\text{or}~b\leq a[/tex]

And of course we used axioms and definitions to define [itex]\mathbb{N}[/itex] and to define the order-relation. But I understand that you take these for granted?
evagelos
#13
Jun7-11, 05:45 PM
P: 308
Quote Quote by micromass View Post
Not exactly! The converse of [itex]x_n< L[/itex] is not always [itex]x_n\geq L[/itex]. We have used here that the order relation is total. I.e. that for every a and b, we have

[tex]a\leq b~\text{or}~b\leq a[/tex]

And of course we used axioms and definitions to define [itex]\mathbb{N}[/itex] and to define the order-relation. But I understand that you take these for granted?

Sorry, i should have mentioned the law of trichotomy (a>b or a=b or a<b),which is the only axiom involved in the proof that:

[tex]\neg\forall n[n\in N\Longrightarrow x_{n}\geq L][/tex] [tex]\Longrightarrow[/tex][tex]\exists n_{1}[n_{1}\in N\wedge x_{n_{1}}< L][/tex]

the rest is logic
micromass
#14
Jun7-11, 05:48 PM
Mentor
micromass's Avatar
P: 18,017
Maybe you can give a list (or reference) to all the axioms you're using? To me, trichotomy isn't really an axiom...
evagelos
#15
Jun7-11, 10:13 PM
P: 308
Quote Quote by micromass View Post
Maybe you can give a list (or reference) to all the axioms you're using? To me, trichotomy isn't really an axiom...
It is not important whether we consider the trichotomy law is an axiom or a theorem .

It is important that we are using it in proving:



[tex]\neg\forall n[n\in N\Longrightarrow x_{n}\geq L][/tex] [tex]\Longrightarrow[/tex][tex]\exists n_{1}[n_{1}\in N\wedge x_{n_{1}}< L][/tex]

Unless you can produce a proof of the above where you can use different theorems ,axioms e.t.c
micromass
#16
Jun7-11, 10:22 PM
Mentor
micromass's Avatar
P: 18,017
Quote Quote by evagelos View Post
It is not important whether we consider the trichotomy law is an axiom or a theorem .
It is important in the sense that I need to know where to begin. What axioms can you assume, what theorems can you assume, that's what I need to know first before beginning this question...
asalmon14
#17
Jun8-11, 11:28 PM
P: 2
How are we defining "x is decreasing?" (I'm sorry, I don't know how to write LaTex)

Say we define a decreasing function
Xn
such that
n1>n0 iff xn1 < xn0.

So then
~ n > infinity => ~ xn < xinfinity
xinfinity = L,
so it seems it would follow that xn > L.

I guess we need to assume infinity (or maybe that for all natural numbers N, there is a successor of N which is greater), the concept of a limit and a definition of a decreasing function.

And of course predicate logic for the substitution that occurs to move from the definition of the decreasing function to substituting infinity for n0.
micromass
#18
Jun9-11, 07:38 AM
Mentor
micromass's Avatar
P: 18,017
You can't do that. Infinity is not a part of the natural numbers, so you can't talk about infinity and things like xinfinity...


Register to reply

Related Discussions
Could someone take a look at an analysis proof? Calculus & Beyond Homework 9
IVP proof (analysis) Calculus & Beyond Homework 1
Analysis Proof Calculus 3
Analysis - What are they doing in this proof ? Calculus & Beyond Homework 4
Analysis proof Calculus & Beyond Homework 1