Mathematics desk
< May 21	<< Apr \| May \| Jun >>	Current desk >

Welcome to the Wikipedia Mathematics Reference Desk Archives
The page you are currently viewing is an archive page. While you can leave answers for any questions shown below, please ask new questions on one of the current reference desk pages.

May 22[edit]

All Solutions for F'(x) = F(x+a).[edit]

We know, for instance, that $\exp '(x)=\exp(x+0),$ $\sin 'x=\sin {\bigg (}x+{\frac {\pi }{2}}{\bigg )},$ and $\cos 'x=\cos {\bigg (}x+{\frac {\pi }{2}}{\bigg )}.$ But since all these are connected through Euler's formula, I was wondering whether there might not be other (non-exponential) functions with the same general property. — 79.118.167.195 (talk) 08:46, 22 May 2014 (UTC)[reply]

If a is fixed and not zero, then you can recover a solution by imposing basically aribitrary data in the form of a smooth function in [0,a] (with appropriate compatibility conditions at the end points) just by integration, I believe. Sławomir Biały (talk) 11:35, 22 May 2014 (UTC)[reply]

Note 'smooth' here has a precise definition, see smooth function. Dmcq (talk) 12:17, 22 May 2014 (UTC)[reply]

F(x) = 0 is a trivial solution for any a. Gandalf61 (talk) 12:26, 22 May 2014 (UTC)[reply]

Note that it suffices to consider a = 0 and a = 1. Count Iblis (talk) 14:45, 22 May 2014 (UTC)[reply]

You also have

\exp '(x)=\exp(x+2in\pi ),

for all integer n (including zero). The same function but more values for a. You can obviously add 2n

π

to

π

/2 in the trigonometric relations too. This suggests there are no more functions as the trigonometric and exponential functions satisfy F'(x) = F(x+a) for so many values.--JohnBlackburne^words_deeds 15:04, 22 May 2014 (UTC)[reply]

I don't really know anything about this equation, but it is a delay differential equation (linear with discrete delay), and that article contains a little bit of general information. —Kusma (t·c) 20:57, 22 May 2014 (UTC)[reply]