comp.lang.idl-pvwave archive: archive » lamda default value in CURVEFIT function

Home » Public Forums » archive » lamda default value in CURVEFIT function

Show: Today's Messages :: Show Polls :: Message Navigator
E-mail to friend

lamda default value in CURVEFIT function [message #86639]

Mon, 25 November 2013 07:05

fd_luni
Messages: 66
Registered: January 2013

Member

Hi

The CURVEFIT function uses the Levenberg–Marquardt algorithm which is based on the equation:

(TRANSPOSE(J)J+ lamda*diag(TRANSPOSE(J)J)(P1-P0)=TRANSPOSE(J)E

J-Jacobian matrix
P0-initial parameter
E-residual

I was wondering if the default value of lamda is given by the keyword TOL and the default value is the 10^-3.

With Thanks
Mar

Report message to a moderator

Re: lamda default value in CURVEFIT function [message #86659 is a reply to message #86639]

Tue, 26 November 2013 08:08

Craig Markwardt
Messages: 1869
Registered: November 1996

Senior Member

On Monday, November 25, 2013 10:05:52 AM UTC-5, fd_...@mail.com wrote:
> Hi
>
>
>
> The CURVEFIT function uses the Levenberg–Marquardt algorithm which is based on the equation:
>
>
>
> (TRANSPOSE(J)J+ lamda*diag(TRANSPOSE(J)J)(P1-P0)=TRANSPOSE(J)E
...
>
> I was wondering if the default value of lamda is given by the keyword TOL and the default value is the 10^-3.

No. The Levenberg-Marquardt parameter (lambda) is adjusted automatically by the algorithm during the fit process.

CM

Report message to a moderator

Previous Topic:	save problems
Next Topic:	"infinite" nested for

-=] Back to Top [=-

[ Syndicate this forum (XML) ] [

] [

]

Current Time: Wed Oct 08 15:07:19 PDT 2025

Total time taken to generate the page: 0.00454 seconds