Given a sequence ((xn,yn)) is R2 .prove that the following,
1. ((xn,yn)) is convergent implies ((xn,yn)) is bounded.
2. If ((xn,yn)) is a bounded sequence the ((xn,yn)) has a convergent subsequence
3. ((xn,yn)) is convergent if and only if ((xn,yn)) is bounded and every convergent subsequence of ((xn,yn)) has the same limit.
4. ((xn,yn)) is caught if and only if ((xn,yn)) is convergent
Solution:
1.
Statement: Every convergent sequence is a bounded sequence, that is the set "\\left\\{x_{n}: n \\in \\mathbb{N}\\right\\}" is bounded.
Proof : Suppose a sequence "\\left(x_{n}\\right)" converges to x. Then, for "\\epsilon=1" , there exist N such that
"\\left|x_{n}-x\\right| \\leq 1 \\text { for all } n \\geq N"
This implies "\\left|x_{n}\\right| \\leq|x|+1" for all "n \\geq N" . If we let
"M=\\max \\left\\{\\left|x_{1}\\right|,\\left|x_{2}\\right|, \\ldots,\\left|x_{N-1}\\right|\\right\\}"
then "\\left|x_{n}\\right| \\leq M+|x|+1" for all n. Hence "\\left(x_{n}\\right)" is a bounded sequence.
2.
Statement: Every bounded sequence contains a convergent subsequence.
Proof. Let "\\left(a_{n}\\right)" be a bounded sequence. Then there exists some M>0 such that "\\left|a_{n}\\right| \\leq M" for all "n \\in \\mathbb{N}" . This implies that "a_{n} \\in[-M, M]" for all "n \\in \\mathbb{N}" .
- We bisect the interval [-M, M] into two closed intervals of equal length. One of these intervals must contain infinitely many terms of the sequence "\\left(a_{n}\\right)" . Let "I_{1}" be that interval, and let "a_{n_{1}}" be any point in "I_{1}" .
- Next we bisect "I_{1}" into two closed intervals, and note that one of these intervals must contain infinitely many terms of the sequence "\\left(a_{n}\\right)". Let "I_{2}" be that interval, and choose "a_{n_{2}}" inside this interval which satisfies "n_{2} \\geq n_{1}"
- In general, we bisect "I_{k-1}" into two closed intervals, one which must contain infinitely many terms of "\\left(a_{n}\\right)" . Let "I_{k}" be this closed interval, and choose "a_{n_{k}} \\in I_{k}" such that "n_{k}>n_{k-1}" . Therefore we have obtained a subsequence "\\left(a_{n_{1}}, a_{n_{2}}, \\ldots\\right)" of "\\left(a_{n}\\right)" and a sequence of nested intervals
"I_{1} \\supseteq I_{2} \\supseteq I_{3} \\supseteq \\ldots"
By the Nested Interval Property (i), the intersection "\\bigcap_{n=1}^{\\infty} I_{n}" is nonempty, and therefore contains some element
We claim that "\\left(a_{n_{k}}\\right)" converges to "x" . Let "\\epsilon>0" be arbitrary. Then the length of the interval "I_{k}" is "M\\left(\\frac{1}{2}\\right)^{k-1}" . By (i), the sequence "\\left(\\frac{1}{2}^{k-1}\\right)" converges to 0. Therefore, by the Algebraic Limit Theorem, the sequence "\\left(M\\left(\\frac{1}{2}\\right)^{k-1}\\right)" converges to 0 as well. Hence, there exists some "N \\in \\mathbb{N}" such that if "k \\geq N" , then "\\left|M\\left(\\frac{1}{2}\\right)^{k-1}\\right|<\\epsilon" But then "a_{n_{k}}, x \\in I_{k}" , hence "\\left|a_{n_{k}}-x\\right|<\\epsilon" . Therefore "\\left(a_{n_{k}}\\right)" converges to "x" .
3.
Statement: Let "\\left\\{x_{n}\\right\\}" be a bounded sequence all of whose convergent proper subsequences converge to "\\ell" . Prove that "\\left\\{x_{n}\\right\\}" converges to "\\ell" .
Proof: A bounded sequence of real numbers must have at least one convergent subsequence. All of those converge to "\\ell" , and there is at least one of those; call it "\\left(x_{n_{k}}\\right)_{k=1}^{\\infty}" . So for every "\\varepsilon>0" , all except finitely many terms of "\\left(x_{n_{k}}\\right)_{k=1}^{\\infty}" are in the interval with endpoints "\\ell \\pm \\varepsilon" . If infinitely many terms of the original sequence lie outside that small interval, then that is a bounded sequence, which therefore has at least one convergent subsequence. That subsequence converges to some point outside that interval small interval. But that contradicts the hypothesis.
4.
Statement: Cauchy Convergence Criterion: A sequence \left(x_{n}\right) is Cauchy if and only if it is convergent.
Proof: Suppose "\\left(x_{n}\\right)" is a convergent sequence, and "\\lim \\left(x_{n}\\right)=x" . Let "\\epsilon>0" We can find "N \\in \\mathbb{N}" such that for all "n \\geq N,\\left|x_{n}-x\\right|<\\epsilon \/ 2" . Therefore, by the triangle inequality, for all "m, n \\geq \\mathbb{N},\\left|x_{m}-x_{n}\\right| \\leq\\left|x_{m}-x\\right|+\\left|x-x_{n}\\right|<\n\n\\epsilon \/ 2+\\epsilon \/ 2=\\epsilon" . So "(x_n\u200b)" is Cauchy.
Conversely, suppose "\\left(x_{n}\\right)" is Cauchy. Let "\\epsilon>0" . By a result proved in class, "\\left(x_{n}\\right)" is bounded. By Bolzano-Weierstrass, it has a convergent subsequence "\\left(x_{n_{k}}\\right)" with "\\lim \\left(x_{n_{k}}\\right)=x" for some "x" . We can find "K \\in \\mathbb{N}" such that for all "k \\geq K,\\left|x_{n_{k}}-x\\right|<\\epsilon \/ 2" . We can also find M such that for all "m, n \\geq M, \\left|x_{m}-x_{n}\\right|<\\epsilon \/ 2" . Let "N=\\sup \\{K, M\\}" . Then since "n_{k} \\geq k" for all k, if "k \\geq N" , we have that k, "n_{k} \\geq M\\ and\\ n_{k} \\geq K" . Therefore, for all "k \\geq N \\left|x_{n}-x\\right| \\leq\\left|x_{n}-x_{n_{k}}\\right|+\\left|x_{n_{k}}-x\\right|<\\epsilon \/ 2+\\epsilon \/ 2=\\epsilon" by the Triangle inequality.
Therefore, "\\left(x_{n}\\right)" is Cauchy.
Comments
Leave a comment