Re: FunctionMinimum

From: Lorenzo Moneta <Lorenzo.Moneta_at_cern.ch>
Date: Tue, 16 Jun 2009 09:29:24 +0200


Hi,

On Jun 15, 2009, at 10:07 PM, Patrick Dupre wrote:

> I made more investigations, in fact it can take a step size of 0.01
> if the
> parameter is not too high. It looks like that it does not like to
> have a step size less than 2e-6. How can I by pass this limitation ?

why do you need smaller step sizes ? You can change this with the SetPrecision method, but you will then get numerical precision problems when calculating the derivatives.

   Best Regards

Lorenzo

>
> Thank.
>
>> On Mon, 15 Jun 2009, Patrick Dupre wrote:
>>
>>> Hello,
>>> I am using FunctionMinimum and fMinimizer.Minimize,
>>> I set the parameter and initial error with the Add function.
>>> I need to set the init_error at 0.01 but it does not take it.
>> Sorry for the confusion, the right term is step size
>>> 0.1 ik OK, when I set 0.01 it looks likes that it takes 0, in
>>> really:
>>> 0.016 ~ 1/60 !!
>>> How can I fix this behavior ?
>>> Thank
>>
>>
>
> --
> ---
> =
> =
> =
> =
> ======================================================================
> Patrick DUPRÉ | |
> Department of Chemistry | | Phone: (44)-
> (0)-1904-434384
> The University of York | | Fax: (44)-
> (0)-1904-432516
> Heslington | |
> York YO10 5DD United Kingdom | | email: pd520_at_york.ac.uk
> =
> =
> =
> =
> ======================================================================
Received on Tue Jun 16 2009 - 09:29:27 CEST

This archive was generated by hypermail 2.2.0 : Tue Jun 16 2009 - 11:50:03 CEST