Let $\Omega$ be an intervall and let $\gamma\in\Omega$. $\lim_{x\rightarrow \gamma}f(x)=L$ is equivalent to the statement bellow.

$$\forall \varepsilon>0\exists \delta:\forall x \in \Omega: 0<\mid x-\gamma\mid<\delta \Rightarrow \mid f(x)-L \mid < \varepsilon$$

I'm working on an assignment where I have to use the above definition of a $\lim_{x\rightarrow \gamma}f(x)=L$ to show that a particular function converges. I've come up with two proofs thus far but I am not certain that the second one is correct since I have allowed $\Omega$ to depend on $\varepsilon$. Is this allowed?

I will give you an example of what I did. At a certain point in the proof I find that $\mid f(x)-L \mid = C\mid x-\gamma\mid ^2$, $C>0$. So if $0<\mid x-\gamma\mid<\sqrt{\frac{\varepsilon}{C}}$ we have that $\mid f(x)-L \mid = C\mid x-\gamma\mid ^2 < C\left(\sqrt{\frac{\varepsilon}{C}}\right)^2=\varepsilon$. If I let $\Omega = \left(\gamma - \sqrt{\frac{\varepsilon}{C}}, \gamma + \sqrt{\frac{\varepsilon}{C}}\right)$ then I can let $\delta = \sqrt{\frac{\varepsilon}{C}}$ correct? Am I allowed to do this or did I mess up?


No you cannot change $\Omega$; but there is no need to anyway. You've already solved it. You can choose $\delta = \sqrt{\frac{\varepsilon}{C}}$ regardless of what $\Omega$ is. It doesn't matter if your $\Omega$ is contained in $(\gamma - \delta, \gamma + \delta)$ or if $(\gamma - \delta, \gamma + \delta)$ is contained in $\Omega$, or neither. To prove the limit is $L$ you must prove that for every $x$ which is in $\Omega$ and in $(\gamma - \delta, \gamma + \delta)$, we have $|f(x)-L|< \epsilon$. It doesn't matter if some points are in $\Omega$ but not in $(\gamma - \delta, \gamma + \delta)$ or vice versa.

Let me know if you have any questions.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.