bugGNU Octave - Bugs: bug #37508, fminsearch misses in deconvolving...

 
 

bug #37508: fminsearch misses in deconvolving 2 gaussian peaks

Submitted by:  None
Submitted on:  Fri 05 Oct 2012 02:55:14 PM UTC  
 
Category: Octave Forge PackageSeverity: 3 - Normal
Priority: 5 - NormalItem Group: Matlab Compatibility
Status: FixedAssigned to: None
Originator Name: CFBOriginator Email: -unavailable-
Open/Closed: ClosedRelease: 3.6.1
Operating System: GNU/Linux

Add a New Comment (Rich MarkupRich Markup):
   

You are not logged in

Please log in, so followups can be emailed to you.

 

Fri 14 Feb 2014 02:52:30 AM UTC, comment #2:

The fminsearch is now a function in Octave itself. I've run the attached exp2 script in Octave 3.8.0 and the results (a1m, m1m, s1m) match the input parameters (a1, m1, s1), so I think this one is fixed now.

If I've misunderstood the problem, or if you still have problems with fminsearch in Octave 3.8.0, please comment or submit a new bug report.

Mike Miller <mtmiller>
Project Member
Fri 05 Oct 2012 03:03:18 PM UTC, comment #1:

This looks like a problem with an Octave Forge package, not a bug
in Octave. The correct place to report this problem is the
Octave Forge mailing list, octave-dev@lists.sourceforge.net.

A copy of this update to the bug report is being sent to that
list so that they are aware of it. Please follow up there.

Jordi GutiƩrrez Hermoso <jordigh>
Project Administrator
Fri 05 Oct 2012 02:55:14 PM UTC, original submission:

1st report from my part. I hope to post in the right place.

I am trying to write code to deconvolve superimposed peaks.
I start by setting up a simulation of 2 gaussian peaks and superimposing them (instead of importing real chromatographic data - best to use "clean" model first, right?). Then I try to deconvolve by minimizing residuals.

Main program is exp2.m, which uses functions gaussian.m and fitexp2.m (see attachments).

Using Matlab at work, it works very nicely. I get the exact same values back that I put in (at 4 decimals for all 6 parameters, 3 from each peak). In GNU Octave, it's a mess.

I haven't hacked "Matlab" code in approx 10 years, so maybe there are better approaches to solve the problem (please enlighten me with any good ideas). However, I would expect them to produce the same results...

Again, I hope I'm doing this the right way. Sorry if it's the wrong place or any other error. I'm relatively new to Linux and supernew to Octave. I have checked for similar post as best I could.

Thanks,
CFB

Anonymous

 

(Note: upload size limit is set to 16384 kB, after insertion of the required escape characters.)

Attach File(s):
   
   
Comment:
   

Attached Files
file #26703:  exp2.m added by None (644B - text/x-objcsrc)
file #26704:  gaussian.m added by None (160B - text/x-objcsrc)
file #26705:  fitexp2.m added by None (191B - text/x-objcsrc)

 

Depends on the following items: None found

Items that depend on this one: None found

 

Carbon-Copy List
  • -unavailable- added by mtmiller (Posted a comment)
  • -unavailable- added by jwe (Updated the item)
  • -unavailable- added by rik5 (Updated the item)
  • -unavailable- added by jordigh (Posted a comment)
  • -unavailable- added by jordigh
  • -unavailable- added by None (Submitted the item)
  •  

    Do you think this task is very important?
    If so, you can click here to add your encouragement to it.
    This task has 0 encouragements so far.

    Only project members can vote.

     

    Please enter the title of George Orwell's famous dystopian book (it's a date):

     

     

    Follow 9 latest changes.

    Date Changed By Updated Field Previous Value => Replaced By
    Fri 14 Feb 2014 02:52:30 AM UTCmtmillerStatusNone=>Fixed
      Open/ClosedOpen=>Closed
    Thu 16 May 2013 08:09:22 PM UTCjweCategoryNone=>Octave Forge Package
      StatusOctave Forge=>None
    Fri 12 Oct 2012 03:50:36 AM UTCrik5StatusNone=>Octave Forge
    Fri 05 Oct 2012 03:03:18 PM UTCjordighCarbon-Copy-=>Added "octave forge list"
    Fri 05 Oct 2012 02:55:14 PM UTCNoneAttached File-=>Added exp2.m, #26703
      Attached File-=>Added gaussian.m, #26704
      Attached File-=>Added fitexp2.m, #26705

    Back to the top


    Powered by Savane 3.1-cleanup