comp.lang.idl-pvwave archive
Messages from Usenet group comp.lang.idl-pvwave, compiled by Paulo Penteado

Home » Public Forums » archive » IDL inverse matrix problem??
Show: Today's Messages :: Show Polls :: Message Navigator
E-mail to friend 
Return to the default flat view Create a new topic Submit Reply
Re: IDL inverse matrix problem?? [message #90070 is a reply to message #90065] Thu, 22 January 2015 13:40 Go to previous messageGo to previous message
Craig Markwardt is currently offline  Craig Markwardt
Messages: 1869
Registered: November 1996
Senior Member
On Thursday, January 22, 2015 at 2:24:06 PM UTC-5, Amin Farhang wrote:

> Am I doing something wrong?
>

I think you are expecting too much from the numerical precision available. Your matrix has a huge range of singular values (factor of 10 billion), so that's going to cause some undesirable round-off errors. The matrices are correct, to within the round-off error of the solution.

If this is very important to you, I think you need to read a lot more about linear algebra. Computing the inverse directly is usually the worst thing you can do. The standard practice is to factor the matrix instead, and solve your problem indirectly. Numerical Recipes has a good amount of beginner discussion on this. Then you can graduate to Golub and van Loan, /Matrix Computations/. Singular Value Decomposition might tell you more about the structure of your matrix.

Craig
[Message index]
 
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Previous Topic: HASH assignment
Next Topic: saving an output from the find.pro script

-=] Back to Top [=-
[ Syndicate this forum (XML) ] [ RSS ] [ PDF ]

Current Time: Fri Oct 10 14:55:23 PDT 2025

Total time taken to generate the page: 0.72139 seconds