[Fityk-users] Smoothing of data

568 views
Skip to first unread message

DEVEL Michel

unread,
Jul 1, 2009, 11:48:31 AM7/1/09
to fityk...@lists.sourceforge.net
Dear Marcyn Wojdir and fityk users,

I have discovered fityk recently and tested it under Windows.
Thanks a lot for taking the pain to make a clean graphical user interface !

I would need such a software to deal with X-ray data of very basic
spectra (W, Ni and Mo with 5 to 7 peaks) for student's lab sessions.
Unfortunately our X-ray setup is quite old and the data is very noisy so
that it should be smoothed before doing the actual peak finding. We
currently have our students using Origin(TM) to do the whole procedure,
but I would like to switch to a free equivalent.
I could not find in the help of fityk nor in the (more or less recent)
archives of this list, something concerning the smoothing of data for
example with a FFT filter or with a simple moving average.
Could someone tell me whether this can already be done by a not too
complicated script ?
If not, could that be put on the TODO list ?

Thanks in advance for your replies !

Marcin Wojdyr

unread,
Jul 2, 2009, 12:07:14 PM7/2/09
to fityk...@lists.sourceforge.net
On Wed, Jul 1, 2009 at 17:48, DEVEL Michel<Michel...@ens2m.fr> wrote:

> Unfortunately our X-ray setup is quite old and the data is very noisy so
> that it should be smoothed before doing the actual peak finding. We

I'm not sure if the data should be smoothed. Fitting algorithms should
give better results with original data.

Anyway, there are no smoothing algorithms implemented, but it is
possible to calculate a simple average, e.g.:
Y[1..-1] = (y[n-1]+y[n]+y[n+1])/3

I'll add smoothing to the TODO list, but since I don't have much time
to work on fityk, it's hard to say when and if at all the items from
TODO list will be implemented.

Marcin

--
Marcin Wojdyr | http://www.unipress.waw.pl/~wojdyr/

DEVEL Michel

unread,
Jul 3, 2009, 8:13:19 AM7/3/09
to fityk...@lists.sourceforge.net
@Marcin: Now that I have had some time to use fityk on my test cases, I do confirm that fityk can indeed find good peaks (previously approximately positioned) in spite of the noise. However, I still have to experiment to use efficiently the base line suppression.
Thanks again for the nice job!

@Jackson: No Pb to let you know if someone else gives me some script about data filtering.
I would be grateful if you could mail me yours  when you will have completed it.
--
Cordialement,

Michel DEVEL

Jackson Read

unread,
Jul 2, 2009, 10:05:43 AM7/2/09
to fityk...@lists.sourceforge.net
Hi Michel,

I am working on this I will let you know how it goes. Please let me know
if you find anything from another user.

Jackson Read

Jackson Read

unread,
Jul 6, 2009, 2:24:12 PM7/6/09
to fityk...@lists.sourceforge.net

By extension you can also implement a simple Savitsky-Golay smoothing w/the same method

--- On Thu, 7/2/09, Marcin Wojdyr <woj...@gmail.com> wrote:

> ------------------------------------------------------------------------------
> _______________________________________________
> Fityk-users mailing list
> Fityk...@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/fityk-users
>



Jackson Read

unread,
Jul 6, 2009, 11:11:15 AM7/6/09
to fityk...@lists.sourceforge.net
Has anybody compiled fityk successfully on windows using mingw? I am
trying to increase the performance of fityk on windows and I have heard
the mingw does not have the overhead of the cygwin gnu compiled program.
I am also having very long build times for fityk (> 5 mins) so I am
hoping that ming will allow me to develop faster. Thanks in advance for
any info

Jackson

Reply all
Reply to author
Forward
0 new messages