error in .peaks export

156 views
Skip to first unread message

Brian

unread,
Mar 9, 2012, 3:40:22 PM3/9/12
to fityk-users
Is there any way to get the .peaks export to also show the uncertainty
in the parameters?

Currently, to get the uncertainties, I have to copy the info output
and then, since I'm using a Voigt function, calculate the FWHM and do
some fun error propagation to get its uncertainty. Also, I've used
the intensity version of the fit, and have just found interest in also
knowing its Area. I have a lot of files fit, and I don't want to go
through and redo them all with VoigtA and then re-export all the fit
info. How do I calculate the Voigt Area from the a0,a1,a2,a3 of the
intensity version? Then I could also figure out the area's
uncertainty. I've been trying to work it out, but I must be making a
math error somewhere and can't get the right number. I could then
just add these calculations to the little script I made to do all my
FWHM calculations.

Thanks

Marcin Wojdyr

unread,
Mar 12, 2012, 9:43:20 PM3/12/12
to fityk...@googlegroups.com
On Fri, Mar 9, 2012 at 20:40, Brian <waiting...@gmail.com> wrote:
> Is there any way to get the .peaks export to also show the uncertainty
> in the parameters?

Sorry for late reply.
You can use "info peaks_err" instead of "info peaks",
or in the GUI open Fit > Info and copy content.
(Last year, I don't remember in which version, I've added confidence
intervals to the latter. I've read somewhere that they can be more
meaningful than estimates of standard errors).

> Currently, to get the uncertainties, I have to copy the info output
> and then, since I'm using a Voigt function, calculate the FWHM and do
> some fun error propagation to get its uncertainty.  Also, I've used
> the intensity version of the fit, and have just found interest in also
> knowing its Area.  I have a lot of files fit, and I don't want to go
> through and redo them all with VoigtA and then re-export all the fit
> info.  How do I calculate the Voigt Area from the a0,a1,a2,a3 of the
> intensity version?

Function voigt(a,b) (see
http://fityk.nieto.pl/data.html#data-point-transformations for
definition) should help.

Marcin

manon...@gmail.com

unread,
Jan 22, 2014, 8:42:17 AM1/22/14
to fityk...@googlegroups.com
Hi Marcin,

I read your message with great interest. You talked about confidence level for uncertainties. Is it possible to export the peak parameters values with uncertainties by choosing a certain confidence level (95%)? To go further: is there any way to obtain a table directly, such as with "info peaks", i.e. a column for 'center', a column for 'area'....?

Thanks in advance for your answer,

Manon

Marcin Wojdyr

unread,
Feb 2, 2014, 11:23:17 AM2/2/14
to fityk...@googlegroups.com
Hi Manon

> I read your message with great interest. You talked about confidence level
> for uncertainties. Is it possible to export the peak parameters values with
> uncertainties by choosing a certain confidence level (95%)? To go further:

Hi Manon,

there is a command "info confidence" for this
info confidence 95 > my-file

> is there any way to obtain a table directly, such as with "info peaks", i.e.
> a column for 'center', a column for 'area'....?

I'm afraid not (not directly).

It's not a bad idea to add it. Actually it would be good to have a
better, more flexible alternative to "info peaks" - there are many
requests regarding this command. But it's not clear to me how exactly
it should look like.

Marcin
Reply all
Reply to author
Forward
0 new messages