bugGNU Octave - Bugs: bug #65182, [octave forge] (image) "test...

 
 

bug #65182: [octave forge] (image) "test stretchlim" fails

Submitter:  Dmitri A. Sergatskov <dasergatskov>
Submitted:  Fri 19 Jan 2024 07:07:22 PM UTC
   
 
Category:  Octave Forge Package Severity:  3 - Normal
Priority:  5 - Normal Item Group:  Unexpected Error or Warning
Status:  None Assigned to:  None
Originator Name:  Open/Closed:  * Open
Release:  * 9.0.90 Operating System:  * GNU/Linux
Fixed Release:  None Planned Release:  None
* Mandatory Fields

Add a New Comment Rich Markup
   

Wed 14 Feb 2024 11:38:51 AM UTC, comment #1: 

Thanks for reporting this behavior, Dmitri.

There is currently already an eps tolerance added to this failing BIST. It is in the image repository already, but not yet in the published image package release:

   https://hg.code.sf.net/p/octave/image/rev/849afbab4913
   (in the context of the bug #63643)

Therefore, I suggest this bug report can be closed as FIXED (or similar).

(Maybe it's time for a new image package release, this year...)


Hartmut <hardy>
Fri 19 Jan 2024 07:07:22 PM UTC, original submission:  

"test stretchlim" on octave 9.0.1 (compiled with AVX).

octave:18> test stretchlim
***** test
 assert (stretchlim ([(.05:.05:1) (2:4)], 0.2), [0.25; 0.95], eps)
 assert (stretchlim ([(.05:.05:1) (2:5)], 0.2), [0.25; 1])
 assert (stretchlim ([(.05:.05:1) (2:6)], 0.2), [0.3; 1])
 assert (stretchlim ([(.05:.05:1) (2:7)], 0.2), [0.3; 1])
!!!!! test failed
ASSERT errors for:  assert (stretchlim ([(.05:.05:1), (2:6)], 0.2),[0.3; 1])

  Location  |  Observed  |  Expected  |  Reason
    (1)          0.3          0.3        Abs err 5.5511e-17 exceeds tol 0 by 6e-17
octave:19> format bit
octave:20> 0.3
ans = 0011111111010011001100110011001100110011001100110011001100110011
octave:21> stretchlim ([(.05:.05:1), (2:6)], 0.2)
ans =

  0011111111010011001100110011001100110011001100110011001100110100
  0011111111110000000000000000000000000000000000000000000000000000

octave:22> stretchlim ([(.05:.05:1), (2:7)], 0.2)
ans =

  0011111111010011001100110011001100110011001100110011001100110100
  0011111111110000000000000000000000000000000000000000000000000000


This "fraction-of-EPS" roundoff error seems typical when AVX instructions are in use.
perhaps the function is OK and the test precision just needs to be relaxed by EPS.

Sincerely,
Dmitri.
--

Dmitri A. Sergatskov <dasergatskov>

 

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

Attach Files:
   
   
Comment:
   

No files currently attached

 

Depends on the following items: None found

Items that depend on this one: None found

 

Carbon-Copy List
  • -email is unavailable- added by hardy (Posted a comment)
  • -email is unavailable- added by dasergatskov (Submitted the item)
  •  

    There are 0 votes so far. Votes easily highlight which items people would like to see resolved in priority, independently of the priority of the item set by tracker managers.

    Only group members can vote.

     

    No changes have been made to this item

    Back to the top

    Powered by Savane 3.12.
    Corresponding source code