| Re: problem inverting matrix [message #33429 is a reply to message #33427] |
Mon, 06 January 2003 07:49   |
Lars Schmidt
Messages: 3
Registered: January 2003
|
Junior Member |
|
|
"David Fanning" <david@dfanning.com> schrieb im Newsbeitrag
news:MPG.188349dfdf6cdc2f989a8d@news.frii.com...
> Lars Schmidt (lars@itandesign.com) writes:
>
>> I desperately hope someone of you has got an idea.
>> I try to invert a matrix using "invert" of course with floating-point
>> values.
>> I get like explosion of values although the status variable I used with
>> "invert" did not return any error and I also used the "/double" keyword.
>> Has anybody an idea or experience with that. Are these numerical
problems of
>> the gaussian ellemination??
>
> When these numbers "exploded", did anyone get hurt? There
> is too much of this going on in the world right now. :-(
No, nobody was. It was actually an attempt of a (as you might suspect)
peace-loving German student to explain the problem.
Thanks for your impression that it is a numerical problem.
By the time all of my variables are DOUBLE.
So, with this rather complex Penrose-Moore part it is not a good thing to
publish in a newsgroup as it is hard to describe.
Thanks for your help.
>
> Did the numbers explode in the direction of 0?
> That is, do these numbers look like some deranged
> person's attempt to represent a very small number?
> If so, it may just be your computer playing tricks on
> you. Computers have a very difficult time representing
> something as small as zero.
>
> You might be interested in this article:
>
> http://www.dfanning.com/math_tips/sky_is_falling.html
>
> Cheers,
>
> David
>
> --
> David W. Fanning, Ph.D.
> Fanning Software Consulting, Inc.
> Phone: 970-221-0438, E-mail: david@dfanning.com
> Coyote's Guide to IDL Programming: http://www.dfanning.com/
> Toll-Free IDL Book Orders: 1-888-461-0155
|
|
|
|
comp.lang.idl-pvwave archive
Members
Search
Help
Login
Home
