chrisidefix / visvis

Automatically exported from code.google.com/p/visvis
Other
0 stars 0 forks source link

np.float128 does not exist in my numpy version, hence revision 9cb8394a404e causes major problems #70

Closed GoogleCodeExporter closed 9 years ago

GoogleCodeExporter commented 9 years ago
__What steps will reproduce the problem?__
1. Revision  9cb8394a404e tests for np.float128
2. In my Python distribution, import numpy; numpy.float128 raises an 
AttributeError, and so does visvis (e.g. when using the LopesApp)

__What is the expected output? What do you see instead?__
Solution: check if numpy.float128 exists before testing if something is an 
instance of it.

This may be a problem occurring in all windows installations (see 
http://stackoverflow.com/questions/5350342/can-i-set-float128-as-the-standard-fl
oat-array-in-numpy, second comment on first post)

__What version of the product are you using? On what operating system?__

I tried three Python installations (all on the same Windows PC). One of them 
lacks only float128, the two others lack both float96 and float128.

Operating system:
- Windows 7 Enterprise Service Pack 1 (64 bits operating system) on Intel 
Core2Duo;

+ Configuration 1 (numpy.float128 does not exist):
  - Python 3.2.1 (default, Jul 10 2011, 21:51:15) on Windows (32 bits)
    This is the IEP interpreter with integrated event loop for PYSIDE.
  - numpy.version.version = '1.6.1'

+ Configuration 2 (numpy.float96 and numpy.float128 don't exist):
  - Python 3.2.3 (default, Apr 11 2012, 07:15:24) on Windows (32 bits).
    This is the IEP interpreter with integrated event loop for PYSIDE
  - numpy.version.full_version = '1.6.2'
  - (this is the configuration from Pyzo)

+ Configuration 3 (numpy.float96 and numpy.float128 don't exist):
  - Python 3.3.0 (v3.3.0:bd8afb90ebf2, Sep 29 2012, 10:55:48) on Windows (32 bits).
    This is the IEP interpreter with integrated event loop for PYSIDE.
  - numpy.version.full_version = '1.7.0'

Original issue reported on code.google.com by gijsvano...@gmail.com on 11 Mar 2013 at 11:12

GoogleCodeExporter commented 9 years ago
Darn... I was afraid of this and tried to google what numpy version has 
float128. Since I could not find anything, I assumed its always there ...

Original comment by almar.klein@gmail.com on 11 Mar 2013 at 12:36

GoogleCodeExporter commented 9 years ago
This issue was closed by revision 30ab843f6682.

Original comment by almar.klein@gmail.com on 11 Mar 2013 at 12:45