what is this?
Pettersen, Bjorn S
BjornPettersen at fairisaac.com
Mon Nov 3 18:56:07 EST 2003
> From: Hari Pulapaka [mailto:pulapaka at ligo.caltech.edu]
please don't top-post.
> Hi,
>
> Most newbies to python, have probably experienced this one time or
> another. And posted it to the newsgroups, where the Python
> gurus, will say, No foo thats how floating points are represented,
> read the tutorial and understand floating point representation, etc...
All you need to remember is that floating point is not exact, and any
attempt at pretending it is _will_ cause other _silent_ problems.
> But call it whatever, 3.4 != 3.3999999999999999 and I feel that
That's the point, there is no 3.4 <wink>.
> somewhere some change should be made so that the
> representation is better.
... by that I take it you mean better == less precise? (If you've read
the afore-mentioned manual, I'm genuinely curious why you're still
thinking this is a good idea?)
> I saw some discussion earlier where, someone provided a function for
> repr so that the presentation of floating is better and it
> looked like (to me) that it had a solution for this problem.
Use str() or %f with appropriate modifiers for presentation.
-- bjorn
More information about the Python-list
mailing list