[proFit-list] Calculation of errors in fits

[Saskia Bosma]

Thanks for your quick and helpful answers ! Actually, I had looked for the
manual in the support page on the website, but not on the download page.

Best regards,
Saskia Bosma

2010/8/4 Arthur Rowe <arthur.rowe at nottingham.ac.uk>

> Absolutely right, Kurt. But - Saskia - pro Fit also makes it exceptionally
> easy to do a more complete error analysis, once you have found (usually by
> repeat analyses) what the errors are in the data points. You simply select
> that option in the Fit. Using 500 iterations you get a parameter
> distribution (for n parameters) and after binning and plotting out you can
> have a good idea as to whether usual assumptions (gaussian) make sense or
> not. Without going into the particular science we were working on, you can
> see by inspection of some of such plots that it is perfectly possible for
> distributions are VERY non-Gaussian indeed (
> http://dx.doi.org/10.1002/mabi.201000065).
>
> The ease with which you can do this sort of thing is a major plus-feature
> of pro Fit.
>
> Kind regards
>
> Arthur
>
>
> *******************************************************************************
> Arthur J Rowe
> Professor of Biomolecular Technology / Director NCMH Business Centre
> School of Biosciences
> University of Nottingham
> Sutton Bonington
> Leics LE12 5RD
>
> TEL:  0115 9516156
> FAX:  0115 0516157
>
> *******************************************************************************
>
>
>
> On Aug 4, 2010, at 04:22, pro Fit Support wrote:
>
>  Dear Saskia
>>
>> For a description of that topic, see page 115 (bottom) and 116 (top) of
>> the manual (http://www.quansoft.com/manual/proFit_61_Handbook.pdf)
>>
>> In short, the errors are in that case only derived from the covariance
>> matrix and chi-squared. In addition, the manual says: "If you need to work
>> with the covariance matrix, and in general for any serious statistical
>> analysis of a fit, you must define errors for your data points!"
>>
>> Best regards
>>
>> Kurt Sutter
>>
>>
>>
>> On 3. Aug 2010, at 17:14, kaasis wrote:
>>
>>  Hello all,
>>>
>>> I am new to profit, and I would like to know how the errors on fit
>>> parameters are calculated. It does not seem to need a Delta_x ? Is it only
>>> derived from the covariance matrix ? I did not find any indication in the
>>> help files or in previous messages on this list.
>>>
>>> Best regards,
>>> Saskia
>>> _______________________________________________
>>> proFit-list mailing list
>>> proFit-list at quantum-soft.com
>>> http://quantum-soft.com/mailman/listinfo/profit-list_quantum-soft.com
>>>
>>
>>
>> _______________________________________________
>> proFit-list mailing list
>> proFit-list at quantum-soft.com
>> http://quantum-soft.com/mailman/listinfo/profit-list_quantum-soft.com
>>
>>
>>
>
> _______________________________________________
> proFit-list mailing list
> proFit-list at quantum-soft.com
> http://quantum-soft.com/mailman/listinfo/profit-list_quantum-soft.com
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://quantum-soft.com/pipermail/profit-list_quantum-soft.com/attachments/20100804/ecc8230e/attachment.html>