MathGroup Archive 1998

[Date Index] [Thread Index] [Author Index]

Search the Archive

RE: Eigenvalues



Running your example on a NT Pentium 150 box took 4.6 seconds. This is 
altogether fast and slow: fast if you realize the answer is pages long,
slow if you have to perform the same calculation with wider matrixes.

Mind the following facts:
- Timing will give you the cpu time used by Mathematica, which can be 
useful if you work on a workstation that runs other programs; - Timing
does not include the time needed by the FrontEnd to display your 
results, ans sometimes the display can take a certain time, in this
case 4  s for the kernel and more than 20 for the FrontEnd. -
Performing symbolic computations is slower than numerical computations,
and humerical computations on large matrixes have never been very fast!

Hope this helps,


----------------------------------------------- Jean-Marie THOMAS
Conseil et Audit en Ingenierie de Calcul www.cybercable.tm.fr/~jmthomas
------------------------------------------------


-----Message d'origine-----
De:	Raya Khanin [SMTP:raya@maths.ed.ac.uk] Date:	jeudi 12 mars 1998
07:35
A:	mathgroup@smc.vnet.net
Objet:	[mg11476] Eigenvalues

Hi,

I have a problem of calculating eigenvalues  for a simple 4X4 matrix. I
am working on Sun workstation, and it takes minutes and minutes for
Mathematica to  perform Eigenvalues[] operation for a matrix like this

Clear[a, b, c, d, e, d, f, g];
testmatrix = {{a, b, c, d}, {b, c, d, e}, {c, d, e, f}, {d, e, f, g}};
Eigenvalues[testmatrix];

I will need to work with much larger (symbolic) matrices with not that
simple entries. Has anyone had an experience of working with symbolic
matrices? Would it be better doing this in MatLab?

Any suggestions will be appreciated!

thanks a lot.

Raya Khanin





  • Prev by Date: Re: ImplicitPlot Problem
  • Next by Date: TELLING PROGRAM THAT A VARIABLE IS AN INTEGER
  • Prev by thread: Re: Re: Eigenvalues
  • Next by thread: Re: Eigenvalues