0
$\begingroup$

$$f(x)=\lim_{n\to\infty}\left(\frac{n^n}{n!}\prod_{r=1}^n\frac{x^2+\frac{n^2}{r^2}}{x^3+\frac{n^3}{r^3}}\right)^{\frac xn}$$What is the monotonicity of $f(x),x\gt0$

I tried in the following way [METHOD-$1$]:

$$f(x)=\lim_{n\to\infty}\left(\frac{n^n}{n!}\prod_{r=1}^n\frac{x^2+\frac{n^2}{r^2}}{x^3+\frac{n^3}{r^3}}\right)^{\frac xn} \implies \lim_{n\to\infty}\left(\frac{1}{n!}\prod_{r=1}^n\frac{\left(\frac xn\right)^2+\frac{1}{r^2}}{\left(\frac xn\right)^3+\frac{1}{r^3}}\right)^{\frac xn}\qquad......... (1)$$ Now for any finite $x$, $$\lim_{n\to\infty} \frac xn \to 0$$ This gives, $f(x)= \left( \left(\frac 1{n!}\right)^2 \over n!.\left(\frac 1{n!}\right)^3\right)^0 \implies (1)^0 = 1$ $\implies f(x)=1$

But here is another approach [METHOD-$2$]:

Continuing from $(1)$, Taking natural log both sides gives; [let $f(x)=y$] $$ln(y) = \lim_{n\to\infty}\frac xn \sum_{r=1}^n ln\left(\frac{\left(\frac xn\right)^2+\frac{1}{r^2}}{\left(\left(\frac xn\right)^3+\frac{1}{r^3}\right)r}\right) \implies \lim_{n\to\infty}ln\left(\frac xn \sum_{r=1}^n \frac{\left(\frac {xr}n\right)^2+1}{\left(\frac {xr}n\right)^3+1}\right)$$ Taking, $\frac rn = t$ and $\frac 1n=dt$; $$\implies ln(y)=\int_0^1ln\left(1+(xt)^2\over1+(xt)^3\right).(xdt)$$ Now $xt=k$ and $xdt=dk$, $$ln(y)=\int_0^xln\left(1+(k)^2\over1+(k)^3\right).(dk)$$ Differentiating both sides, $$\frac{y'}y=ln\left(1+(k)^2\over1+(k)^3\right)$$

Now since $y\ge0$;

$y'\ge0$ for $0\lt x\le1$

$y'\le0$ for $x\ge1$

So $f(x)$ is increasing for $0\lt x\lt1$ and $f(x)$ is decreasing for $x\gt1$

The answer in the 2 cases do not match.

And the answer in METHOD-$2$ happens to be the correct answer.

I need help to find what is wrong with METHOD-$1$

$\endgroup$
4
  • 1
    $\begingroup$ In the method 1, you take the limit in a part of the expression, then compute the product, and then take the limit again. You can't do that. Try to compute the limit of $(1 + x/n)^n$ when $n \rightarrow +\infty$ and you will see that you can't just replace the $x/n$ by $0$, it doesn't give you the right expression. $\endgroup$
    – Cactus
    Commented Apr 5 at 15:30
  • $\begingroup$ @Cactus your reasoning doesn't seem much satisfactory to me and also the example stated by you is of the form $1^{\infty}$ which is undefined so it is not applicable while where i have applied it, it is of the form $1^0$ so i think it should be applicable $\endgroup$ Commented Apr 5 at 15:57
  • $\begingroup$ I don't know if this will help or not but $$f(x) = \frac{e^{x+ 2\arctan (x)} }{ e^{x \int_0^1 \frac{1}{1+x^3t^3} dt }} \cdot \left( \frac{1+x^2}{1+x^3} \right)^x $$ A closed form of $ \int_0^1 \frac{1}{1+x^3t^3} dt $ exists but it is very messy to write. $\endgroup$
    – pie
    Commented Apr 5 at 17:20
  • 2
    $\begingroup$ The problem with METHOD1 is your simplification of $\infty^0$ to $1$, which is not true. Check $\lim_{x\to \infty} (a^x)^\frac{1}{1+x}$ for example. $\endgroup$ Commented Apr 8 at 11:50

0

You must log in to answer this question.

Browse other questions tagged .