Re: [Visualpython-users] inconsistent arange() behavior

[Bruce S. <Bru...@nc...>]

It's nothing more than the fact that floating-point numbers can be 
represented only approximately in a computer. A simple example from 
ordinary pencil-and-paper calculations is that 1/3 + 1/3 + 1/3 = 1, but 
0.33+0.33+0.33 = .99. If you want arange to produce 0, "1/3", "2/3", but 
not "1", you should write something like arange(0.0, 1.0-.1, 1.0/3.0). 
More generally, if you want to go from xi to xf in floating-point 
increments dx, and not include xf, write arange(xi, xf-0.5*dx, dx). 
Another solution is something like this, using integers:

while n in range(ni, nf, dn):
	x = scale*n # where scale is a floating-point scale factor
	....

Bruce Sherwood

Frédéric Mantegazza wrote:
> On dimanche 30 mars 2008, Joe Heafner wrote:
> 
>> Okay this is driving me crazy. Typing arange(0.1,0.4,0.1) should give
>> [0.1, 0.2, 0.3] but instead it gives [0.1, 0.2, 0.3, 0.4]. Typing
>> arange(0.1,0.5, 0.1) should give [0.1, 0.2, 0.3, 0.4] and it indeed
>> does. Another example that doesn't give the "correct" result is
>> arange(0.2,0.8,0.2), which gives [ 0.2, 0.4, 0.6, 0.8]. Is this a bug in
>> the arange() code? I've tinkered with this for two days and there seems
>> to be no way to predict when arange() will behave as documented. I have
>> my students working on a small assignment that uses arange() and I don't
>> want this to affect their work. Is there some detail I'm overlooking?
> 
>     arange([start,] stop[, step,], dtype=None)
> 
>     For integer arguments, just like range() except it returns an array
>     whose type can be specified by the keyword argument dtype.  If dtype
>     is not specified, the type of the result is deduced from the type of
>     the arguments.
> 
>     For floating point arguments, the length of the result is ceil((stop -
>     start)/step).  This rule may result in the last element of the result
>     being greater than stop.
> 
> I'm afraid you will have to check the result, and correct it. But I agree, 
> this is a strange behaviour!
>