bugGNU Octave - Bugs: bug #56569, Inconsistent error with larger...

 
 

bug #56569: Inconsistent error with larger matrices

Submitter:  None
Submitted:  Mon 01 Jul 2019 01:59:24 PM UTC
   
 
Category:  Octave Function Severity:  3 - Normal
Priority:  5 - Normal Item Group:  Incorrect Result
Status:  Invalid / Not an Octave Bug Assigned to:  None
Originator Name:  Mark van Rossum Originator Email:  -email is unavailable-
Open/Closed:  * Closed Release:  * 4.4.1
Operating System:  * GNU/Linux Fixed Release:  None
Planned Release:  None
* Mandatory Fields

Add a New Comment Rich Markup
   

Jump to the original submission

Tue 02 Jul 2019 09:08:56 AM UTC, comment #6: 

Thanks Sebastien,
It turned out to be indeed an MKL incompatibility.

After removing those libraries, things are working correctly.


Mark van Rossum <cosine>
Tue 02 Jul 2019 08:54:39 AM UTC, comment #5: 

In particular, it could be the documented incompatibility between Octave and MKL.

See Debian bug 921207: https://bugs.debian.org/921207

Sébastien Villemot <svillemot>
Tue 02 Jul 2019 12:58:16 AM UTC, comment #4: 

It could be due to a broken blas library.
You can try running the affected octave with different blas e.g.:

reference blas


LD_PRELOAD=/usr/lib64/libblas.so.3 octave


atlas



LD_PRELOAD=/usr/lib64/atlas/libtatlas.so octave


etc...

Dmitri.
--

Dmitri A. Sergatskov <dasergatskov>
Mon 01 Jul 2019 09:19:44 PM UTC, comment #3: 

(I am the original reporter).

This is on a Debian system with Octave 4.4.1-5.

I have now compiled 5.1 and there it is ok as well.

It is a very odd bug, any tips to further narrow it down appreciated.

Mark van Rossum <cosine>
Mon 01 Jul 2019 03:32:00 PM UTC, comment #2: 

I also can't reproduce this with either Octave 4.4 or 5.1.

Mike Miller <mtmiller>
Group Member
Mon 01 Jul 2019 03:20:42 PM UTC, comment #1: 

ubuntu 18.04 and octave 5.1.1
I get
N =  1000
M =  500
ans =  500
ans = 0

Doug Stewart <dastew>
Mon 01 Jul 2019 01:59:24 PM UTC, original submission:  

The script below calculates a simple matrix product
p=M.a

For the larger matrices the output fluctuates across the elements in vector p.

Works fine on Matlab, and Octave 4.2.2 on another computer.

------
# this goes ok. All elements in the  vector p are the same.
N=100
M=50
m=ones(N,M);
a=ones(M,1);
p=m*a;
mean(p)
var(p) # should be zero

# this goes wrong. The elements in the  vector p vary.
N=1000
M=500
m=ones(N,M);
a=ones(M,1);
p=m*a;
mean(p)
var(p) # should be zero

Anonymous

 

(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 svillemot (Posted a comment)
  • -email is unavailable- added by dasergatskov (Posted a comment)
  • -email is unavailable- added by cosine (Posted a comment)
  • -email is unavailable- added by dastew (Posted a comment)
  • -email is unavailable- added by None (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.

     

    Follow 3 latest changes.

    Date Changed by Updated Field Previous Value => Replaced by
    2019-07-02 mtmiller StatusWorks For Me Invalid / Not an Octave Bug
        Open/ClosedOpen Closed
    2019-07-01 mtmiller StatusNone Works For Me

    Back to the top

    Powered by Savane 3.13-3230.
    Corresponding source code