Convergence

Convergence

• 1.
  INTRODUCTION TO THECONVERGENCE OF SEQUENCES BECKY LYTLE Abstract. In this paper, we discuss the basic ideas involved in sequences and convergence. We start by defining sequences and follow by explaining convergence and divergence, bounded sequences, continuity, and subsequences. Relevant theorems, such as the Bolzano-Weierstrass theorem, will be given and we will apply each concept to a variety of exercises. Contents 1. Introduction to Sequences 1 2. Limit of a Sequence 2 3. Divergence and Bounded Sequences 4 4. Continuity 5 5. Subsequences and the Bolzano-Weierstrass Theorem 5 References 7 1. Introduction to Sequences Definition 1.1. A sequence is a function whose domain is N and whose codomain is R. Given a function f: N → R, f(n) is the nth term in the sequence. Example 1.2. The first example of a sequence is xn = 1 n . In this case, our function f is defined as f(n) = 1 n . As a listed sequence of numbers, this would look like the following: (1.3) 1, 1 2 , 1 3 , 1 4 , 1 5 , 1 6 , 1 7 ... Another example of a sequence is xn = 5n , which would look like the following: (1.4) (5, 25, 125, 625...) We know these are both valid examples of sequences because they are infinite lists of real numbers and hence can be regarded as functions with domain N. Example 1.5. The following would not be an examples of sequences: (1.6) (1, 2, 3) (1.7) (500, 200, 550, 10000) Date: July 12, 2015. 1
• 2.
  2 BECKY LYTLE Weknow that these are not examples of sequences because they are finite lists of real numbers. 2. Limit of a Sequence A limit describes how a sequence xn behaves “eventually” as n gets very large, in a sense that we make explicit below. Definition 2.1. A sequence of real numbers converges to a real number a if, for every positive number , there exists an N ∈ N such that for all n ≥ N, |an - a| < . We call such an a the limit of the sequence and write limn→∞ an = a. Proposition 1. The sequence 1 n converges to zero. Proof. Let > 0. We choose N ∈ N such that N > 1 . Such a choice is always possible by the Archimedean property. To verify that this choice of N is appropri- ate, let n ∈ N satisfy n ≥ N. Then, n ≥ N implies n > 1 which is equal to 1 n = | 1 n −0 | < , proving that 1 n converges to zero by the definition of convergence. Proposition 2. An example of a sequence that does not converge is the following: (2.2) (1, −1, 1, −1, ...) If a sequence does not converge, it is said to diverge, which we will explain later in the paper, along with the explanation of why the above sequence does not converge. Proposition 3. If xn ≤ yn ≤ zn for all n ∈ N and limn→∞ xn=limn→∞ zn=l, then limn→∞ yn=l too. Proof. Let > 0. We want to show there exists an N such that for all n > N, | yn − l |< . We know that xn → l. Therefore, there exists an N1 such that for all n > N1, | xn − l |< . Also, we know that zn → l. Therefore, there exists an N2 such that for all n > N2, | zn − l |< . Let N = max(N1, N2) and n > N. Then, n > N1 so | xn − l |< . Also, n > N2 so | zn − l |< . We want to show that | yn − l |< . This is equivalent to showing that both yn − l < and l − yn < . We know that yn ≤ zn, so yn − l ≤ zn − l < . Also, we know that yn ≥ xn, so l − yn ≤ l − xn < . Theorem 2.3 (Algebraic Limit Theorem). Let limn→∞ an = a and limn→∞ bn = b. Then, (i) limn→∞ can = ca for all c ∈ R (ii) limn→∞(an + bn) = a+b (iii) limn→∞(anbn) = ab (iv) limn→∞(an/bn) = a/b provided b = 0 Example 2.4. If (xn) → 2, then ((2xn - 1)/3) → 1. Proof. First, we will start with the information given in the example: xn → 2. Next, we simply use the fact that (4 3 )(3 2 ) = 2. (2.5) xn → 4 3 3 2
• 3.
  INTRODUCTION TO THECONVERGENCE OF SEQUENCES 3 Now, let an = xn, and let a = (4 3 )(3 2 ), and let c = (2 3 ). From the Algebraic Limit Theorem, we know that can → ca. Then, (2 3 )(xn) → (2 3 )(4 3 )(3 2 ), which is equal to the following: (2.6) 2xn 3 → 4 3 The next step follows from the fact that 4 3 = 1 + 1 3 . (2.7) 2xn 3 → 1 + 1 3 Let 2xn 3 = an, let (1+1 3 ) = a, let bn = (−1 3 , −1 3 , ...), and let b = −1 3 . Then, by the Algebraic Limit Theorem, we know that an + bn → a + b. Therefore, we know that 2xn 3 + −1 3 → (1+1 3 ) + −1 3 , which is equal to the following: (2.8) 2xn 3 − 1 3 → 1 This last step follows because 2xn 3 - 1 3 = 2xn−1 3 . (2.9) 2xn − 1 3 → 1 Therefore, using the Algebraic Limit Theorem, we have shown that if (xn) → 2, then ((2xn - 1)/3) → 1. Example 2.10. The following sequence converges to the proposed limit (2.11) lim 2n + 1 5n + 4 = 2 5 Proof. Let 2n 5n+4 be an, let 1 5n+4 be bn and let 2n+1 5n+4 be cn, and cn = an + bn. By Theorem 2.3, we know that lim(cn) = lim (an+bn) = lim(an)+lim(bn). We must therefore determine what lim(an) and lim(bn) are. First, we will show that lim( 1 5n+4 ) = 0. Let > 0. By the Archimedean principle, there exists an N ∈ N such that N > 1/ . Then, for n > N, 1 5n+4 < 1 5N+4 < 1/N < . Therefore, the limit of 1 5n+4 is zero. Then, because lim(cn) = lim(an+bn), lim(cn) = lim(an + 0) = lim(an). We will therefore find the limit of an in order to prove lim(2n+1 5n+4 ) = 2 5 . We now want to show that lim( 2n 5n+4 ) = 2 5 . Let > 0. By the Archimedean Principle, there exists an N such that 1/ < N. Let n > N. We then want to show the following: (2.12) 2n 5n + 4 − 2 5 < Then, (2.13) 2n 5n + 4 − 2 5 = −8 5(5n + 4)
• 4.
  4 BECKY LYTLE Wehave to check the following: (2.14) −8 5(5n + 4) < (2.15) 8 5(5n + 4) < We know that the inequality −8 5(5n+4) < is true for every value of n because n > N > 1/ and . Therefore we only need to show that the inequality 8 5(5n+4) < is true. Using the fact that N > 1/ , we can say the following: (2.16) 8 5(5n + 4) < 8 5(5(1/ ) + 4) = 8 25 + 20 Then, 8 25+20 < 8 25 < . Therefore, 8 5(5n+4) < . Example 2.17. Let xn ≥ 0. If (xn) → 0, then ( √ xn) → 0. Proof. First, we have to prove that lim( √ xn) exists. We know that xn is decreasing but is greater than or equal to 0 for all values of n. The square root of a positive number is also positive. Therefore, √ xn ≥ 0. Also, note that if 0< a < b, then 0< √ a< √ b. So if xn is decreasing, then so is √ xn. Therefore, lim( √ xn) exists. Next, we must prove that ( √ xn) → 0. Let lim(xn) = lim(( √ xn)( √ xn)) = 0. By the Algebraic Limit Theorem, we know that if lim(an) = a and lim(bn) = b then lim((an)(bn)) = ab. By this theorem, lim(( √ xn)( √ xn)) = lim( √ xn)lim( √ xn) = 0. Thus, (lim( √ xn))2 =0. This implies that lim( √ xn)=0. 3. Divergence and Bounded Sequences Definition 3.1. A sequence that does not have a limit or in other words, does not converge, is said to be divergent. Example 3.2. Recall proposition 2, which says that the following sequence does not converge: (3.3) (1, −1, 1, −1...) Later in this paper, we will give a concise proof of this fact. Contrast this with the following sequence, which we have seen (3.4) 1, 1 2 , 1 3 , 1 4 , 1 5 , 1 6 , 1 7 ... This converges to zero, as we proved earlier in this paper. However, these se- quences do have something in common. They are both bounded. Definition 3.5. A sequence (xn) is bounded if there exists a number M>0 such that | xn | ≤ M for all n ∈ N. Geometrically, this means we can find an interval [−M,M] that contains every term in the sequence (xn).
• 5.
  INTRODUCTION TO THECONVERGENCE OF SEQUENCES 5 Example 3.6. Given the sequence xn = (1, 2, 1, 2, 1, 2 ...), we can see that the interval [1, 2] contains every term in xn. This sequence is therefore a bounded sequence. Example 3.7. Given the sequence xn = (10, 100, 1000, 10000, ...), we can see that there is no real number that serves as an upper bound because lim(xn) is infinity. Therefore, there does not exist any interval that contains every term in the sequence xn, and xn is not a bounded sequence. Theorem 3.8. Every convergent sequence is bounded. Example 3.9. Theorem being illustrated: Let xn = n+1 n , which is the following sequence: (3.10) 2 1 , 3 2 , 4 3 , 5 4 ... We know this converges to 1 and can verify this using the same logic used in the proof under the definition of convergence showing that 1 n converges to zero. Therefore, as n becomes very large, xn approaches 1, but is never equal to 1. By the above theorem, we know that this sequence is bounded because it is convergent. We can see that xn is a decreasing sequence, so the x1 is the largest value of the sequence and is the “upper bound.” The limit of the sequence, 1, is the lower bound. An interval that contains every term in the sequence xn is (1,2]. 4. Continuity Theorem 4.1. If f: R → R is continuous, xn → x implies f(xn) → f(x) Example 4.2. Theorem being applied: Let f(x) = 3x. This function is continuous. Let lim(xn) = 5. In order words, xn → 5. By the above theorem, this implies that f(xn) → f(5). This is equal to 3xn →(3)(5) which is also equal to 3xn → 15. Therefore, we are able to see what the limit of f(xn) is using this theorem. Example 4.3. Theorem failing when function is non-continuous: Let f(x) be 1 x , a non-continuous function. We know this is non-continuous be- cause there is an asymptote at x=0. Let xn be 1 n . We know this converges to zero based on a previous proof. Let’s see if the continuity theorem fails for a non- continuous function f. The theorem states that f(xn) converges to f(x) if xn → x. We know that xn → 0, so if the theorem works, then f(xn) → f(0). But f(0) = 1 0 which does not exist. Therefore, f(xn) cannot converge to 1/0, and the theorem fails for this non-continuous function. 5. Subsequences and the Bolzano-Weierstrass Theorem Definition 5.1. Let (an) be a sequence of real numbers, and let n1 < n2 < n3 < n4... be an increasing sequence of natural numbers. Then, the sequence (an1 , an2 , an3 , an4 ...) is called a subsequence of (an) and is denoted by (ank ), where k ∈ N indexes the subsequence.
• 6.
  6 BECKY LYTLE Example5.2. Let xn = 1 n = (1, 1 2 , 1 3 , 1 4 ...). Below are two examples of valid subsequences: (5.3) ( 1 3 , 1 6 , 1 9 , 1 12 ...) (5.4) ( 1 20 , 1 200 , 1 2000 ...) Theorem 5.5. Bolzano-Weierstrass Theorem: Every bounded sequence contains a convergent subsequence. Example 5.6. Given a sequence xn = (1,2,3,4,1,2,3,4...), a convergent subsequence can be found. Proof. We know that this sequence is bounded by the interval [1, 4]. By the Bolzano-Weierstrass Theorem, we can say that there indeed exists a convergent subsequence of xn. Just by looking at this sequence, we can see four convergent subsequences: (1,1,1...), (2,2,2...), (3,3,3...), and (4,4,4...). These subsequences con- verge to 1, 2, 3, and 4 respectively. Example 5.7. Given an unbounded sequence xn = (1,2,3,4,5...), a convergent subsequence of xn does not exist Proof. A convergent subsequence does not necessarily exist because this sequence does not satisfy the Bolzano-Weierstrass Theorem. Recall that any subsequence of a sequence is non-repeating and in the order of the original entries of xn. Notice that xn is increasing for all values of n and is divergent, considering the sequence continues until infinity. Therefore, for any subsequence an, the values will be in- creasing toward infinity as well, and the subsequence will also be divergent. Theorem 5.8. Subsequences of a convergent sequence converge to the same limit as the original sequence. Example 5.9. Let us return to the example of a divergent sequence that was given under the definition of divergence. Recall that this sequence, xn, was (1, -1, 1, -1, 1, -1...). One subsequence of xn is (1, 1, 1, ...). This subsequence converges to 1. Another subsequence of xn is (-1, -1, -1, ...). This subsequence converges to -1. Now, we will prove that xn is divergent by contradiction. Assume xn is convergent. Then, by the above theorem, all its subsequences converge to lim(xn), implying that all its subsequences converge to the same value. The two subsequences of xn stated above converge to different values. Therefore, this contradicts our original hypothesis that xn is convergent. We are then able to conclude that xn is divergent. Acknowledgements. I would like to thank my two mentors, Sean Howe and Abby Ward for meeting with me each week for the duration of my stay at the program. Thank you so much for suggesting this topic, explaining everything to me that I was confused about, editing and helping with my paper, and calming me down when I got overwhelmed. I would also like to thank Peter May for organizing this wonderful REU and for giving me the opportunity to participate.
• 7.
  INTRODUCTION TO THECONVERGENCE OF SEQUENCES 7 References [1] Walter Rudin. Principles of Mathematical Analysis. McGraw-Hill Inc. 1976. [2] Stephen Abbott. Understanding Analysis. Springer. 2015.