[SciPy-dev] strange error with win32 build for py21

[Jun Sung]

Dear experts:
 
I am getting very strange error with SciPy-0.2.0-alpha.
All multiplication involving array (array*scalar) coerces fractions into 0's.
I have tracked the error down to fastumath module in scipy_base.
 
For example:
>>> import Numeric
>>> x=Numeric.arange(11.)
>>> x*.5
array([ 0. ,  0.5,  1. ,  1.5,  2. ,  2.5,  3. ,  3.5,  4. ,  4.5,  5. ])
>>> import scipy_base.fastumath
>>> x*.5
array([ 0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.])
 
 
Furthermore, as seen in "scipy_base/limits.py" , when two arrays with float typecodes
are multiplied, Python throws an ArithmeticError: Integer overflow in multiply.
 
Should I bother with the C code or is there an obvious clue I need to get.
Thanks, in advance.
/js


---------------------------------
Do you Yahoo!?
The New Yahoo! Search - Faster. Easier. Bingo.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/scipy-dev/attachments/20030813/d5c5ff1f/attachment.html>