[R] SE from nleqslv

peter dalgaard pdalgd at gmail.com
Tue Mar 20 18:03:11 CET 2012

On Mar 20, 2012, at 15:55 , Berend Hasselman wrote:

> On 20-03-2012, at 15:36, FU-WEN LIANG wrote:
>> On Tue, Mar 20, 2012 at 4:24 AM, Berend Hasselman <bhh at xs4all.nl> wrote:
>>> On 20-03-2012, at 01:01, FU-WEN LIANG wrote:
>>>> Dear R-users,
>>>> I use the "nleqslv" function to get parameter estimates by solving a
>>>> system
>>>> of non-linear equations. But I also need standard error for each of
>>>> estimates. I checked the nleqslv manual but it didn't mention about SE.
>>>> Is there any way to get the SE for each estimate?
>>> nleqslv is for solving a nonlinear system of equations. Only that.
>>> If you provide a system of equations for determining standard errors then
>>> nleqslv might be able to solve that system.
>>> You can use nleqslv to investigate the sensitivity of a solution wrt
>>> changes in parameters.
>>> Berend
>> Thank you very much for your advice, Berend.
>> Would you please give me a hint about "the sensitivity of a solution
>> wrt changes in parameters"? What statistics can we use?
> Suppose you have a system of two equations and this system depends on two parameters A and B.
> You have solved the system for specific values of A and B.
> Then you can vary A and B see how the solution changes.
> How that could or might be translated to SE's I really wouldn't know.
> A measure of the sensitivity could be the (relative change of a norm of the solution) / (relative change of a parameter).

Well, the delta method springs to mind, but it really depends on how and where noise is being injected into the system. All we have been told is that the estimates are obtained as a solution to a nonlinear equation, and that can mean many things. Presumably there are some observations somewhere, with a distribution, etc.

> Berend
> ______________________________________________
> R-help at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.

Peter Dalgaard, Professor,
Center for Statistics, Copenhagen Business School
Solbjerg Plads 3, 2000 Frederiksberg, Denmark
Phone: (+45)38153501
Email: pd.mes at cbs.dk  Priv: PDalgd at gmail.com

More information about the R-help mailing list