On Fri, 12 Dec 2003 Harm_Kirchhoff@[...].jp wrote:

>  I have the following problem:
> 
>  >>> float(8.7)
>  8.6999999999999993
>  >>> float(870)/100
>  8.6999999999999993
>  >>> 8.7 * 100
>  869.99999999999989
>  >>> round(8.7,2)
>  8.6999999999999993
>  >>> math.modf(8.7)
>  (0.69999999999999929, 8.0)
>  >>> int(float(8.7*100))
>  869

This is what int is suppose to do. As lib/built-in-funcs.html states:
"Conversion of floating point numbers to integers truncates (towards
zero)."

better:

[user@host]# wget http://starship.python.net/crew/aahz/FixedPoint.py

[user@host]# python

> >> import FixedPoint
> >> dir(FixedPoint)
['DEFAULT_PRECISION', 'FixedPoint', '__builtins__', '__doc__',
'__file__', '__name__', '__version__', '_mkFP', '_norm', '_parser',
'_roundquotient', '_string2exact', '_tento', '_test']
> >> f = FixedPoint.FixedPoint(8.7)
> >> f
FixedPoint('8.70', 2)
> >> f * 100
FixedPoint('870.00', 2)
> >> int(f * 100)
870

--->  a simple programming-language arithmetic isn't good enough for
Money Application. You will loose some money when you use it (or
gain some ;-).

So you should look out for solutions made for this problem (google for
"python+money"). FixedPoint seems for me like a very good starting
point to write your programm correctly. Perhaps FixedPoint offers
allready everything you need.

But you might also want support different types of commercial
rounding conventions (Shall $4.225 be rounded up or down?). I've heard
some people work on a Money class to solve exactly these type of things
but it is still in progress. There was a python-dev thread not long ago:
http://mail.python.org/pipermail/python-dev/2003-October/038902.html

This was a "prePEP" (which stands for Python Enhancemeent Proposal) as
you notice but it (and the further discussion) might give you some very
valueable informations.


Michael








_______________________________________________
Tutor maillist  -  Tutor@[...].org
http://mail.python.org/mailman/listinfo/tutor