Quantcast
Channel: Active questions tagged real-analysis - Mathematics Stack Exchange
Viewing all articles
Browse latest Browse all 8476

Solve for $m(t)$ in the integral transform $\int_0^1(1-t^n) m(t) dt=\frac{(n+1)^n}{n^{n-1}} $ for $n>0$.

$
0
0

Background

(You can skip this part, but maybe you find it interesting.)

Is $ \displaystyle f_1(x,v) = \sum_{n=0}^{\infty} \frac{x^n}{(n!)^v } > 0 $ for all real $x$ and $0<v<1$ ?

Lets start simple and take the case $v = \frac{1}{2}$ and notice the inverse ratio of Taylor coefficients is $(a_{n+1}/a_n)^{-1} = \sqrt n$. Probably the simplest way to prove positivity is now to write

$$ \int_{0}^{1} (1-t^n)\left(\log\frac{1}{t}\right)^{-3/2} \, \frac{dt}{t} = c\sqrt{n}$$

with some fixed positive $c$. Notice that the integral converges and the integrand is positive, and make the change of variable $t^n\to t$. We conclude,by the additive property and the identity $x^m t^m = (xt)^m$, that

$$ \int_{0}^{1} (f(x)-f(xt))\left(\log \frac{1}{t} \right)^{-3/2} \, \frac{dt}{t} = cxf(x). $$

If $x$ is the largest zero of $f$ (which must be negative), then by plugging it in, we get $0$ on the right and a negative number on the left, which is a clear contradiction. Thus, crossing the $x$-axis is not possible.

Of course, there is nothing special about $v = \frac{1}{2}$. Any power $v$ between $0$ and $1$ works just as well because the analogue integral still converges. For instance,

\begin{gather*}\int_{0}^{1}(1-t^n)\left(\log \frac{1}{t}\right)^{-5/4} \, \frac{dt}{t} = 4 \Gamma(3/4) n^{1/4} \qquad\text{and} \\\int_{0}^{1} (f(x)-f(xt)) \left(\log \frac{1}{t}\right)^{-5/4} \, \frac{dt}{t} = cxf(x).\end{gather*}

This method cries out for a generalization and a deeper understanding.

Notice also that the limit of $f(x,v)$ as $x \to -\infty$ is zero and the function is strictly increasing. Also, partially because of that, the function is estimated (and proven) to go to zero at rate $O(n^{-v})$ (not so easy to prove?) and to $+\infty$ at rate $O(\exp(x^{1/v}))$.


So this is where the idea came from. There are other ways to prove positivity, but I really like this one. (A conjecture is that all similar questions with confirmation of positivity need to be reduced or be reducible to the above, or be a sequence of compositions that themselves are.)

So, we focus on the integral transform$M(t(n)) = m(t)$ for a given $t(n)$, such that

\begin{equation}\int_{0}^{1} (1-t^n) m(t) \, dt = t(n) \label{int_tf}\tag{*}\end{equation}

is valid for integer $n>0$ with $m(t)$ being positive for $0<t<1$.

However, it seems

$$ \lim_{x \to -\infty} \sum_{n=1}^{+\infty} \frac{x^n}{(n+1)^n} = -1 $$

AND even

$$ 1 + \sum_{n=1}^{+\infty} \frac{x^n}{(n+1)^n} > 0 $$

This made me wonder about the integral transform...


EDIT 1

After some testing it seems more like

$$ \lim_{x \to -\infty} \sum_{n=1}^{+\infty} \frac{x^n}{(n+1)^n} = -1$$

AND

$$0 \leq C + \sum_{n=1}^{+\infty} \frac{x^n}{(n+1)^n} $$

Where $C = 1.036055393...$ is the best possible constant.

This seems to suggest that the integral transform does not exist since the function does not start from the empty product or equivalently

$$1 + \sum_{n=1}^{+\infty} \frac{x^n}{(n+1)^n} > 0$$

Is false. But there has to be a better more direct way to show it right?


Question

Consider the integral transform $M(t(n)) = m(t)$ as in \eqref{int_tf}. I wonder if there is a solution of

$$ M((n+1)^n) = m(t),$$

and if so, what it is? So, we want to find strict positive $m(t)$ such that

$$\int_{0}^{1} (1-t^n) m(t) \, dt= \frac{(n+1)^n}{n^{n-1}}$$

valid for integer $n>0$.

It seems like a reasonable question.And it seems closely related to Laplace and similar transforms or the Bernstein theorem.Yet I'm not sure how to do that.

I think I can restate the problem with Lambert-W functions or so, but then again I doubt that would help much.

(Notice $n$-th derivatives of Lambert-W are $\frac{(-n)^n}{n!}$ and you already see the similarity.)

Any ideas?


EDIT 2 (A similar conjecture)

Not essential but beyond a comment perhaps.

How to prove that for all real $x$ we have:

$$ 0 \leq f(x) = A + \sum_{n=1}^{+\infty} \frac{x^n}{(n+1)^{\frac{n}{2}}} $$

Where $A = 1$ is the best possible constant.

Somewhat related or similar: Conjecture about $ f(x,v) = 1 + \sum_{n=1}^\infty {x^n \over t(v n) } > 0 $ for all real $x$


Viewing all articles
Browse latest Browse all 8476

Trending Articles



<script src="https://jsc.adskeeper.com/r/s/rssing.com.1596347.js" async> </script>