[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: Fitting with leasqr and uncertainties
From: |
Olaf Till |
Subject: |
Re: Fitting with leasqr and uncertainties |
Date: |
Mon, 20 Jan 2014 15:37:29 +0100 |
User-agent: |
Mutt/1.5.21 (2010-09-15) |
On Mon, Jan 20, 2014 at 05:27:37AM -0800, st.michal88 wrote:
> Thank you very much for your reply. I found an information that square root
> of diagonal elements of covariant matrix is approximately the error of
> fitted parameters. To check that I used on the other hand Origin to fit the
> data as it presents the results of calculations as parameter+-uncertainty.
> It was the same as the error from covariance matrix. Is it correct then?
Not necessarily. The variation in a certain linear combination of
parameters could have more influence on a parameter than the variation
only in the parameter (which you get from the diagonal of the
covariance matrix). The applicability of your above suggestion will
depend on the absolute value of the off-diagonal values of the
covariance matrix; for more exact statements you would probably have
to do a principal component analysis.
Olaf
--
public key id EAFE0591, e.g. on x-hkp://pool.sks-keyservers.net
signature.asc
Description: Digital signature
Re: Fitting with leasqr and uncertainties, Juan Pablo Carbajal, 2014/01/20
Re: Fitting with leasqr and uncertainties, CdeMills, 2014/01/20