3
$\begingroup$

The following is a common identity between dilogarithms: $$ \text{Li}_2(y)+\text{Li}_2\left(\frac{y}{y-1}\right)+\frac{1}{2} \log ^2(1-y)=0\quad \text{with}\quad 0<y<1\,. $$

Similarly, with some effort, one can find an identity involving $\text{Li}_3(y)$, $\text{Li}_3\left(\frac{y}{y-1}\right)$ and $\text{Li}_3(1-y)$ $$ \text{Li}_3(1-y)+\text{Li}_3(y)+\text{Li}_3\left(\frac{y}{y-1}\right)-\frac{1}{6} \log ^3(1-y)+\frac{1}{2} \log (y) \log ^2(1-y)-\frac{1}{6} \pi ^2 \log (1-y)-\zeta (3)=0, $$ with $0<y<1$ and the above being proved for instance in Lewin L. Polylogarithms and associated functions (NH, 1981). Do you know if there is an identity that expresses

$$ \text{Li}_4\left(\frac{y}{y-1}\right) $$ as a function of $\text{Li}_4(y)$ and $\text{Li}_4(1-y)$?

$\endgroup$
1
  • $\begingroup$ The short answer is no, unfortunately. $\endgroup$
    – David H
    Dec 19, 2021 at 9:19

1 Answer 1

1
$\begingroup$

After some effort, I was able to find a relation of the sort I (almost) wanted. As anticipated by David H, the relation I really wanted probably does not exist. However, if you are willing to live with a relation that involves more than just $\mathrm{Li}_4(y)$, $\mathrm{Li}_4(1-y)$ and $\mathrm{Li}_4(\frac{y}{1-y})$, then the following is probably the most concise thing one can write down $$ \text{Li}_4(1-y)=\frac{1}{24} \pi ^2 \log ^2(1-y)-\frac{19 \pi ^4}{6480}+\frac{7}{48} \log ^4(1-y)-\frac{1}{6} \log ^3(1-y) \log (y)+\frac{1}{3} \text{Li}_4\left(-\frac{y}{(1-y)^2}\right)+\frac{1}{18} \text{Li}_4\left(-(1-y)^3\right)-\frac{3}{2} \text{Li}_4(-(1-y))-\frac{1}{3} \text{Li}_4\left(-\frac{y}{1-y}\right)+\frac{1}{6} \text{Li}_4\left(\frac{y}{1-y}\right)-\frac{1}{6} \text{Li}_4((1-y) y)+\frac{1}{6} \text{Li}_4\left(\frac{y^2}{(1-y)^2}\right)+\frac{1}{6} \text{Li}_4\left(-\frac{y^2}{1-y}\right)-\frac{1}{18} \text{Li}_4\left(\frac{y^3}{(1-y)^3}\right)\,. $$ One remarkable thing about the above expression is that it tells you immediately what is the singularity structure of $\text{Li}_4(x)$ for $x\sim1$. This identity can be found rather hidden (and with a typo) in Lewin L. Polylogarithms and associated functions (NH, 1981).

$\endgroup$
1
  • $\begingroup$ very interesting identity. $\endgroup$ Jan 27, 2022 at 15:55

You must log in to answer this question.

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