bugGNU Octave - Bugs: bug #49900, cplxpair return error message...

 
 

bug #49900: cplxpair return error message where data seems good

Submitted by:  Henk Borsje <hjborsje>
Submitted on:  Wed 21 Dec 2016 12:23:41 PM UTC  
 
Category: Octave FunctionSeverity: 3 - Normal
Priority: 5 - NormalItem Group: Incorrect Result
Status: DuplicateAssigned to: None
Originator Name: Henk BorsjeOpen/Closed: Closed
Release: 4.2.0Operating System: Microsoft Windows

Add a New Comment(Rich Markup)
   

You are not logged in

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

 

Thu 14 Sep 2017 04:19:34 PM UTC, comment #3:

Henk - if you think there is still a bug in cplxpair, please file a new bug report with a detailed description.

With the example you posted here (or how I understand it)

I get no error, seems to be working fine.

Mike Miller <mtmiller>
Project Administrator
Thu 14 Sep 2017 03:59:48 PM UTC, comment #2:

It looks like there still is a problem with cplxpair(). It seems unrelated to units of tolerance, as someone suggested, it is a different bug.

When the function is entered with:

z = -1.0000 + 1.2246e-16

tol = 2.2204e-10

the same error still appears in the following section of code:

## For each remaining z, place the value and its conjugate at the start of
## the returned list, and remove them from further consideration.

for j = 1:m
p = n - nr(j);
for i = 1:2:p
if (i+1 &gt; p)
error (&quot;cplxpair: could not pair all complex numbers&quot;);
endif

'nr' seems to be an empty sparse matrix and 'p' gets the value of 1 and always has a value less than 'i+1'

The input arguments come from a call to 'butter' and are not made up.

Thank you for your attention.

Henk Borsje

Henk Borsje <hjborsje>
Wed 21 Dec 2016 06:02:12 PM UTC, comment #1:

Thanks for reporting this bug. This is a duplicate of bug #47865, which is not yet fixed. Please continue with any comments or contributions on that report.

Mike Miller <mtmiller>
Project Administrator
Wed 21 Dec 2016 12:23:41 PM UTC, original submission:

With the following input arguments ML version of cplxpair returns valid answers, Octave reports error

x =

-0.309016994374947 + 0.951056516295154i
-0.809016994374947 + 0.587785252292473i
-1.000000000000000 + 0.000000000000000i
-0.809016994374947 - 0.587785252292473i
-0.309016994374948 - 0.951056516295154i

tol =

2.482534375291878e-09

Octave returns error in following lines:

if (v > tol*eps (abs (z(i))))
error ("cplxpair: could not pair all complex numbers");
endif

ML returns:

-0.809016994374947 - 0.587785252292473i
-0.809016994374947 + 0.587785252292473i
-0.309016994374947 - 0.951056516295154i
-0.309016994374947 + 0.951056516295154i
-1.000000000000000

Henk Borsje <hjborsje>

 

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

Attach File(s):
   
   
Comment:
   

No files currently attached

 

Digest:
   bug dependencies.

Items that depend on this one: None found

 

Carbon-Copy List
  • -unavailable- added by mtmiller (Posted a comment)
  • -unavailable- added by hjborsje (Submitted the item)
  • -unavailable- added by hjborsje (cmplpair)
  •  

    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 4 latest changes.

    Date Changed By Updated Field Previous Value => Replaced By
    Wed 21 Dec 2016 06:02:12 PM UTCmtmillerStatusNone=>Duplicate
      Open/ClosedOpen=>Closed
      Dependencies-=>Depends on bugs #47865
    Wed 21 Dec 2016 12:23:41 PM UTChjborsjeCarbon-Copy-=>Added hjborsje

    Back to the top


    Powered by Savane 3.1-cleanup1