[Search for users] [Overall Top Noters] [List of all Conferences] [Download this site]

Conference rusure::math

Title:Mathematics at DEC
Moderator:RUSURE::EDP
Created:Mon Feb 03 1986
Last Modified:Fri Jun 06 1997
Last Successful Update:Fri Jun 06 1997
Number of topics:2083
Total number of notes:14613

220.0. "roots of tan x = x" by HARE::STAN () Sun Feb 10 1985 17:50

Here's a hard problem from the monthly:

Prove that

		---\
		\        1
		 >	---	=  1/10
		/         2
		---/     r
		 r

where the sum is taken over all positive real solutions to  tan x = x .



Reference:

Robert M. Young. Problem 6488. American Mathematical Monthly. 92(1985)148.
T.RTitleUserPersonal
Name
DateLines
220.1SPRITE::OSMANTue Feb 12 1985 17:333
degrees or radians ?  It does matter, right ?  Or perhaps tan(x)=x only
HAS solutions for one of the two units and hence that's part of the
exercise !
220.2HARE::STANTue Feb 12 1985 18:391
Radians is intended.
220.3ADVAX::J_ROTHWed Mar 20 1985 21:5922
Tho I don't see any obvious solution, it may be based on Parseval's
equation.  This states that the energy in a periodic function f(t)
over its period must equal the energy of all its Fourier coefficients.

	 		---
	/     2	       \        2
	| f(t) dt =     >   |c |
	/	       /      n
			---

Say you had f(t) which gave rise to set of Fourier coefficients
which happened to be the roots of tan(r) = r... Then if f(t)^2 could
be integrated in closed form over one period you'd have a proof.  This is one
popular way of summing series (such as 1/n^2) since such series
often result from simple f(t)'s such as triangle waves, sawtooth
waves, etc.

The roots of tan(r) = r also have other suspicious connections; I remember
they're the eigenvalues of a certain boundry value problem so that
might help (i don't remember which one though)

- Jim