[Scipy-tickets] [SciPy] #667: scipy.linalg.toeplitz complex numbers warning

SciPy Trac scipy-tickets@scipy....
Sat Mar 20 20:11:27 CDT 2010


#667: scipy.linalg.toeplitz complex numbers warning
----------------------------------------+-----------------------------------
 Reporter:  wa03                        |       Owner:  somebody    
     Type:  defect                      |      Status:  needs_review
 Priority:  normal                      |   Milestone:  0.8.0       
Component:  scipy.linalg                |     Version:              
 Keywords:  toeplitz conjugate warning  |  
----------------------------------------+-----------------------------------
Changes (by warren):

  * status:  new => needs_review


Comment:

 The problem still exists in the trunk.  A recent discussion on the scipy-
 dev mailing list begins here:

     http://mail.scipy.org/pipermail/scipy-dev/2010-March/013981.html

 I have attached a patch.  It does a little more than fix the bug reported
 in this ticket.  Here's a summary:

 Changes to toeplitz:
 1. Always returns a 2D array, even if either argument is a scalar.
 2. When c is complex and r is not given, r=c.conj() is assumed.  The first
 column
    is always c, not c.conj().
 3. The first element of r is always ignored; the function never prints a
 warning
    if c[0] != r[0].  The docstring explains that r[0] is always ignored.

 A few similar changes were made to hankel (but the default value of r is
 zeros_like(c), not c.conj()).

 I also added a new function, circulant(c), that creates a circulant matrix
 whose first
 column is c.


 The patch changes the API a bit, but it changes behavior that was
 undocumented (and, IMO, unexpected).

-- 
Ticket URL: <http://projects.scipy.org/scipy/ticket/667#comment:2>
SciPy <http://www.scipy.org>
SciPy is open-source software for mathematics, science, and engineering.


More information about the Scipy-tickets mailing list