octave-bug-tracker
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Octave-bug-tracker] [bug #56037] qp exits with an indexing error for a


From: Mike Miller
Subject: [Octave-bug-tracker] [bug #56037] qp exits with an indexing error for a specific input with OpenBLAS
Date: Thu, 4 Apr 2019 15:57:59 -0400 (EDT)
User-agent: Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/73.0.3683.86 Safari/537.36

Follow-up Comment #20, bug #56037 (project octave):

The last part should have read

With OpenBLAS the eigenvectors are


[v, ~] = eig (p)
v =

  -0.22089  -0.02616  -0.00000   0.16667   0.16667   0.83333  -0.27481 
-0.35355
   0.22089   0.02616   0.00000  -0.16667   0.83333   0.16667   0.27481  
0.35355
   0.22089   0.02616   0.00000   0.83333  -0.16667   0.16667   0.27481  
0.35355
   0.22089   0.02616   0.00000  -0.50000  -0.50000   0.50000   0.27481  
0.35355
  -0.11281   0.44394   0.70711  -0.00000  -0.00000   0.00000  -0.40643  
0.35355
  -0.11281   0.44394  -0.70711   0.00000   0.00000  -0.00000  -0.40643  
0.35355
  -0.85950  -0.26116   0.00000  -0.00000  -0.00000   0.00000   0.26086  
0.35355
   0.20156  -0.73136   0.00000   0.00000   0.00000   0.00000  -0.54725  
0.35355



And with netlib BLAS, the eigenvectors are returned as


[v, ~] = eig (p)
v =

  -0.07138   0.16667   0.16667   0.83333   0.22361   0.01465  -0.26399 
-0.35355
   0.07138   0.83333  -0.16667   0.16667  -0.22361  -0.01465   0.26399  
0.35355
   0.07138  -0.16667   0.83333   0.16667  -0.22361  -0.01465   0.26399  
0.35355
   0.07138  -0.50000  -0.50000   0.50000  -0.22361  -0.01465   0.26399  
0.35355
   0.07138  -0.00000  -0.00000   0.00000   0.89443  -0.01465   0.26399  
0.35355
  -0.45433  -0.00000  -0.00000   0.00000  -0.00000  -0.71838  -0.39053  
0.35355
  -0.56908   0.00000   0.00000  -0.00000   0.00000   0.68695  -0.28152  
0.35355
   0.66653   0.00000   0.00000   0.00000   0.00000   0.10468  -0.64791  
0.35355



The first 7 columns correspond to eigenvalues of zero or near to zero, so is
it reasonable that the eigenvectors are not identical? Seems reasonble to me.

I have no idea if there is a bug in how the qp algorithm is implemented when
the system has one or more zero eigenvalues.

    _______________________________________________________

Reply to this item at:

  <https://savannah.gnu.org/bugs/?56037>

_______________________________________________
  Message sent via Savannah
  https://savannah.gnu.org/




reply via email to

[Prev in Thread] Current Thread [Next in Thread]