[Numpy-tickets] [NumPy] #501: "print arr" fails with nan when seterr(all='raise') [test included]

NumPy numpy-tickets@scipy....
Sat May 12 21:30:16 CDT 2007


#501: "print arr" fails with nan when seterr(all='raise') [test included]
-------------------------+--------------------------------------------------
 Reporter:  AndrewStraw  |        Owner:  somebody     
     Type:  defect       |       Status:  new          
 Priority:  normal       |    Milestone:  1.0.3 Release
Component:  numpy.core   |      Version:  none         
 Severity:  normal       |   Resolution:               
 Keywords:               |  
-------------------------+--------------------------------------------------
Comment (by AndrewStraw):

 Using your script I get the following:
 {{{
 Traceback (most recent call last):
   File "ticket501.py", line 7, in ?
     sstr = N.array_str(s)
   File "/home/astraw/py2.4-linux-x86_64/lib/python2.4/site-
 packages/numpy/core/numeric.py", line 473, in array_str
     return array2string(a, max_line_width, precision, suppress_small, ' ',
 "", str)
   File "/home/astraw/py2.4-linux-x86_64/lib/python2.4/site-
 packages/numpy/core/arrayprint.py", line 267, in array2string
     separator, prefix)
   File "/home/astraw/py2.4-linux-x86_64/lib/python2.4/site-
 packages/numpy/core/arrayprint.py", line 179, in _array2string
     format = _floatFormat(data, precision, suppress_small)
   File "/home/astraw/py2.4-linux-x86_64/lib/python2.4/site-
 packages/numpy/core/arrayprint.py", line 353, in _floatFormat
     max_val = max_reduce(non_zero)
 FloatingPointError: invalid value encountered in reduce
 }}}

 So, it's still a problem for me on my Ubuntu Edgy amd64 architecture
 (Intel Core 2 Duo 6600)
 {{{
 Python 2.4.4c1 (#2, Oct 11 2006, 20:00:03)
 [GCC 4.1.2 20060928 (prerelease) (Ubuntu 4.1.1-13ubuntu5)] on linux2
 }}}

 If we check the test in, we'll at least see if others are affected by the
 issue.

-- 
Ticket URL: <http://projects.scipy.org/scipy/numpy/ticket/501#comment:3>
NumPy <http://projects.scipy.org/scipy/numpy>
The fundamental package needed for scientific computing with Python.


More information about the Numpy-tickets mailing list