PEP 260: simplify xrange()

[Tim Peters]

[Dinu Gherman, on extend range/xrange to floats]
> No, I haven't followed the discussion... but you asked
> about an implementation; I'm using the one below.
>
> Dinu
>
>
> def frange(start, end=None, inc=None):
>     "A range function, that does accept float increments..."
>
>     if end == None:
>         end = start + 0.0
>         start = 0.0
>
>     if inc == None:
>         inc = 1.0
>
>     L = [start]
>     while L[-1] + inc < end:
>         L.append(L[-1] + inc)
>
>     return L


>>> print len(frange(0.0, 1.0, 0.1))
11
>>>

That's the kind of "inexplicable" crap we don't want to have to try to
explain <wink>:  repeated addition is a Bad Idea for floats.

You can improve the numeric quality of the results by generating via

    L = []
    while 1:
        next = start + len(L) * inc
        if next >= end:
            break
        L.append(next)

instead (repeated addition accumulates rounding errors without bound; the
alternative suffers no more than 2 rounding errors (total) for each element
produced).  Then it returns 10 elements in the example above.