Very specific quibble re: proving that $g(x)=\sqrt[3]{x}$ is continuous at $c \neq 0$ using $\epsilon-\delta$ arguments

62 Views Asked by At

I'm mulling over an exercise in Abbott's textbook. We want to prove that $\forall \epsilon>0 \exists \delta>0$ such that, if $0<|x-c|<\delta$, then $|g(x)-g(x)|< \epsilon$.

A little bit of algebra confirms that

$$x-c = (x^{1/3} - c^{1/3})(x^{2/3} + c^{1/3}x^{1/3} + c^{2/3})$$

Hence, we can bound our $|g(x)-g(c)|$ appropriately,

$$|x^{1/3} - c^{1/3}| = \frac{|x-c|}{|x^{2/3} + c^{1/3}x^{1/3} + c^{2/3}|} \leq \frac{|x-c|}{c^{2/3}}$$

So choosing $\delta=:\epsilon c^{2/3}$ will do.

This is where I take issue with the answer key. For the original denominator to be greater than $c^{2/3}$, we need to know that $c^{1/3}x^{1/3} > 0$.

But this will only happen if we define $\delta < |c|$, so that transitivity guarantees that

$$|x-c| < |c| \ \Longrightarrow \ c-|c| < x<c+|c| \ \Longrightarrow \ xc>0$$

Shouldn't then $\delta :=\min\{|c|,\epsilon c^{2/3}\}$?

1

There are 1 best solutions below

0
On BEST ANSWER

Yes, you are correct.

Another way to fix the proof would be to say "Without loss of generality, suppose $\epsilon < |c|^{1/3}$."