- #1
lovebear123
- 1
- 0
Spivak "Calculus" Problem 5-23
Spivak "Calculus" Problem 5-23: For your convenience, here is the problem:
Prove that if neither of the following two holds :
1. lim x ->0 f(x) exists and is not 0
2. lim x ->0 |f(x)| = infinity
then there is a function g such that lim x ->0 g(x) does not exist, but lim x ->0 f(x)g(x) does exist. Hint: Consider separately the following two cases : (1) For some e>0, we have |f(x)| > e for all sufficiently small x. (2) For every e>0, there are arbitrarily small x with |f(x)| < e. In the second case, begin by choosing points Xn with |Xn| < 1/n and |f(Xn)| < 1/n.
Note: e means epsilon
My attempt:
Case 1: for some e > 0, as long as x is small enough, we have |f(x)| > e. Let g = 1/f, obviously this does not hold for all e>0 (otherwise
lim x ->0 |f(x)| = infinity). so there is some e1 such that the case title does not hold, there are arbitrarily small x with |f(x)| < e1, for
e' = 1/e1>0, there are arbitrarily small x with |g(x)| = |1/f(x)| > e' = 1/e1. so lim x ->0 g(x) does not exist, and lim x ->0 f(x)g(x) =1
Case 2:for every e>0, no matter how small x is restricted to, there are some x such that |f(x)| < e. Let g = 1/f, for each N>0, according to the assumption, we have for e = 1/N, there are arbitrarily small x with |x| < 1/N and |f(x)| < e = 1/N, so we have arbitrarily small x with |g(x)| > N, so lim x->0 g(x) does not exist and lim x ->0 f(x)g(x) =1
My Questions are:
1. I am totally not sure about my attempt
2. I am confused by the answer given in the answer book (by Spivak), especially concerning Case 2, Spivak writes:
In case 2, choose Xn as in the hint. Define g(x) = 0 for x is not Xn, and g(x) =1 for x= Xn. Then lim x -> 0 g(x) does not exist, but
lim x ->0 f(x)g(x) = 0.
I totally don't understand this, for example, f(x) = sin(1/x), when n =1, X1<1, f(X1)<1, obviously all x satisfying |x| <1 would be suitable Xn (when n = 1), but obviously g(x) =1 for all such x won't make
lim x ->0 f(x)g(x) = 0.
Actually it's a total mess, I don't understand the whole problem at all, I don't know the intention for such two cases...Need help~
I really hope that someone could help me with this, and this is my first experience in this physicsforums, so please... Thank you very much!
Spivak "Calculus" Problem 5-23: For your convenience, here is the problem:
Prove that if neither of the following two holds :
1. lim x ->0 f(x) exists and is not 0
2. lim x ->0 |f(x)| = infinity
then there is a function g such that lim x ->0 g(x) does not exist, but lim x ->0 f(x)g(x) does exist. Hint: Consider separately the following two cases : (1) For some e>0, we have |f(x)| > e for all sufficiently small x. (2) For every e>0, there are arbitrarily small x with |f(x)| < e. In the second case, begin by choosing points Xn with |Xn| < 1/n and |f(Xn)| < 1/n.
Note: e means epsilon
My attempt:
Case 1: for some e > 0, as long as x is small enough, we have |f(x)| > e. Let g = 1/f, obviously this does not hold for all e>0 (otherwise
lim x ->0 |f(x)| = infinity). so there is some e1 such that the case title does not hold, there are arbitrarily small x with |f(x)| < e1, for
e' = 1/e1>0, there are arbitrarily small x with |g(x)| = |1/f(x)| > e' = 1/e1. so lim x ->0 g(x) does not exist, and lim x ->0 f(x)g(x) =1
Case 2:for every e>0, no matter how small x is restricted to, there are some x such that |f(x)| < e. Let g = 1/f, for each N>0, according to the assumption, we have for e = 1/N, there are arbitrarily small x with |x| < 1/N and |f(x)| < e = 1/N, so we have arbitrarily small x with |g(x)| > N, so lim x->0 g(x) does not exist and lim x ->0 f(x)g(x) =1
My Questions are:
1. I am totally not sure about my attempt
2. I am confused by the answer given in the answer book (by Spivak), especially concerning Case 2, Spivak writes:
In case 2, choose Xn as in the hint. Define g(x) = 0 for x is not Xn, and g(x) =1 for x= Xn. Then lim x -> 0 g(x) does not exist, but
lim x ->0 f(x)g(x) = 0.
I totally don't understand this, for example, f(x) = sin(1/x), when n =1, X1<1, f(X1)<1, obviously all x satisfying |x| <1 would be suitable Xn (when n = 1), but obviously g(x) =1 for all such x won't make
lim x ->0 f(x)g(x) = 0.
Actually it's a total mess, I don't understand the whole problem at all, I don't know the intention for such two cases...Need help~
I really hope that someone could help me with this, and this is my first experience in this physicsforums, so please... Thank you very much!
Last edited: