1
$\begingroup$

hello community.

I need some help

given $k = \sum A_q e^{ia_q s}$ where $k$, and $s$ is known. Can $A_q$ s now be expressed in terms of $a_q$ .

any help in that direction will be appreciated.

$\endgroup$
6
  • 1
    $\begingroup$ Is the sum finite or infinite? Is it just one equation or a system? $\endgroup$
    – fedja
    Dec 4, 2011 at 12:47
  • $\begingroup$ finite. one equation. so we have one equation with $n$ unknowns. that is, $a_q$ (or $A_q$) for $ q \in \{1,...,n\}$ $\endgroup$
    – Sean
    Dec 4, 2011 at 14:21
  • $\begingroup$ You should indicate if the $A_q$ and $a_q$ are assumed real (which they likely are). What about $k$ and $s$ ? $\endgroup$ Dec 4, 2011 at 14:25
  • $\begingroup$ in the general case $k$ is complex, $s$ is real. $\endgroup$
    – Sean
    Dec 4, 2011 at 14:39
  • $\begingroup$ Is there something I'm missing? How can you hope to solve one equation for N variables? $\endgroup$ Dec 4, 2011 at 14:57

1 Answer 1

1
$\begingroup$

No, you cannot, at least not at this level of generality. Consider for example $a_q = \frac{\pi}{q}$, $s=1$ and $k=0$. Basically a roots-of-unity situation; then each $A_q$ can be a multiple of the proper coefficient of a cyclotomic polynomial. That does 'solve' the problem, in that you get the $A_q$'s in terms of the $a_q$, up to a single free variable. But of course you can also express a finite linear combination of many roots of unity in the same way, thus getting as many free variables as you wish.

If the $a_q$ have some structure, then you might be able to say something -- thought the problem often reduces to some nasty diophantine equations OR to requiring Schanuel's conjecture to separate things out.

Your question is likely to get closed soon unless you make it a lot more precise and motivated.

$\endgroup$
0

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