-2
$\begingroup$

Question:

Prove using $\epsilon$-$\delta$ definition that if $g$ continous in $a$ and that if $f$ continuous in $g(a)$ then $f\circ g$ is continuous in $a$
Rem: I know that there is a similar question here but this not the same question of mine as I am asking a proof by definition.

My Answer:

1- By assumption we know that because:
(i) $g$ is continuous in $a$ we have by definition that $\forall \epsilon_1>0, \exists \delta_1 > 0 : \forall |x-a|< \delta_1 \Rightarrow |g(x)-g(a)| \epsilon_1$
(ii) $f$ is continuous in $g(a)$ we have by definition that $\forall \epsilon_2>0, \exists \delta_2 > 0 : \forall |y-g(a)|< \delta_2 \Rightarrow |f(y)-f(g(a))| < \epsilon_2$
In order to prove that $f(g(x))$ is continuous by $\epsilon - \delta$ definition at $x=a$, we must show that $\forall \epsilon > 0 , \exists \delta > 0 : \forall |x-a| < \delta \Rightarrow |f(g(x))-f(g(a))| < \epsilon $

2- Given that $f(x)$ is continuous at $g(a)$ then according to (ii) $\forall \epsilon_2>0, \exists \delta_2 > 0 : \forall |y-g(a)|< \delta_2 \Rightarrow |f(y)-f(g(a))| < \epsilon_2$

3- Now because $g(x)$ is continuous at $a$ the definition in (i) is correct forall $\epsilon_1>0$ it is in particular true in the case for $\epsilon_1 = \delta_2 $ so in such a case we have that for $ \epsilon_1 = \delta_2 $ it exists at least one $\delta_2 > 0$ such that for $x$ verifying $|x-a| < \delta_1$ we have that $ |g(x)-g(a)| < \delta_2 = \epsilon_1$

4- Now let fixe one $\epsilon > 0 $
according to "3-" we can choose $\delta = \delta_1 > 0$ such that all the $x$ verifying $|x-a| < \delta_1 $ verify too that $|g(x)-g(a)| < \delta_2$
according to "2-" by the continuity of $f$ at $g(a)$ it implies that $|f(g(x))-f(g(a))|<\epsilon$

And this finish my prove because it is valable for any fixed $\epsilon > 0 $

Q.E.D.

Is it correct? especially part -3 and -4.

$\endgroup$
11
  • $\begingroup$ What does prove by definition mean? $\endgroup$
    – masiewpao
    Commented Mar 27 at 14:16
  • $\begingroup$ @masiewpao thank for your time. You re right I ve edited my question to be more precise. The meaning is by $\epsilon - \delta$ definition. Is it more clear now? $\endgroup$
    – OffHakhol
    Commented Mar 27 at 14:20
  • $\begingroup$ And what exactly is the question? $\endgroup$
    – stange
    Commented Mar 27 at 14:21
  • 2
    $\begingroup$ The solution-verification tag shouldn't be used for questions like "Is my proof correct?". Instead, you should say why you think that your proof might be incorrect. $\endgroup$
    – stange
    Commented Mar 27 at 14:23
  • 1
    $\begingroup$ math.stackexchange.com/questions/527855/… $\endgroup$
    – user408858
    Commented Mar 27 at 18:13

1 Answer 1

0
$\begingroup$

My answer is correct. A similar proove can be found here:
Real Analysis: Continuity of a Composition Function
(Did not know that this answer was posted)

$\endgroup$

Not the answer you're looking for? Browse other questions tagged .