# Bizarre arithmetic results

Discussion in 'Python' started by Terrence Cole, Feb 11, 2010.

1. ### Terrence ColeGuest

Can someone explain to me what python is doing here?

Python 3.1.1 (r311:74480, Feb 3 2010, 13:36:47)
[GCC 4.3.4] on linux2
>>> -0.1 ** 0.1

-0.7943282347242815
>>> a = -0.1; b = 0.1
>>> a ** b

(0.7554510437117542+0.2454609236416552j)
>>> -abs(a ** b)

-0.7943282347242815

Why does the literal version return the signed magnitude and the
variable version return a complex?

Cheers,
Terrence

Terrence Cole, Feb 11, 2010

2. ### Peter OttenGuest

Terrence Cole wrote:

> Can someone explain to me what python is doing here?
>
> Python 3.1.1 (r311:74480, Feb 3 2010, 13:36:47)
> [GCC 4.3.4] on linux2
>>>> -0.1 ** 0.1

> -0.7943282347242815
>>>> a = -0.1; b = 0.1
>>>> a ** b

> (0.7554510437117542+0.2454609236416552j)
>>>> -abs(a ** b)

> -0.7943282347242815
>
> Why does the literal version return the signed magnitude and the
> variable version return a complex?

Operator precedence.

>>> -0.1**0.1

-0.7943282347242815
>>> (-0.1)**0.1

(0.7554510437117542+0.2454609236416552j)

Quoting http://docs.python.org/3.1/reference/expressions.html:

"""
The power operator binds more tightly than unary operators on its left; it
binds less tightly than unary operators on its right.
"""

Peter

Peter Otten, Feb 11, 2010

3. ### Mark DickinsonGuest

On Feb 11, 12:44 am, Terrence Cole <list-
> wrote:
> Can someone explain to me what python is doing here?

> >>> -0.1 ** 0.1

> -0.7943282347242815

Here you're computing -(0.1 ** 0.1). The exponentiation operator
binds more strongly than the negation operator.

> >>> a = -0.1; b = 0.1
> >>> a ** b

> (0.7554510437117542+0.2454609236416552j)

Here you're computing (-0.1) ** 0.1.

--
Mark

Mark Dickinson, Feb 11, 2010
4. ### Jussi PiitulainenGuest

Terrence Cole writes:

> Can someone explain to me what python is doing here?
>
> Python 3.1.1 (r311:74480, Feb 3 2010, 13:36:47)
> [GCC 4.3.4] on linux2
> >>> -0.1 ** 0.1

> -0.7943282347242815
> >>> a = -0.1; b = 0.1
> >>> a ** b

> (0.7554510437117542+0.2454609236416552j)
> >>> -abs(a ** b)

> -0.7943282347242815
>
> Why does the literal version return the signed magnitude and the
> variable version return a complex?

The minus sign is not part of the literal syntax. Python takes the
expression as -(0.1 ** 0.1), the binary operator binding tighter than
the unary.

Try (-0.1) ** 0.1, and try a = 0.1, then -a ** 0.1.

Jussi Piitulainen, Feb 11, 2010
5. ### Grant EdwardsGuest

On 2010-02-11, Terrence Cole <> wrote:
> Can someone explain to me what python is doing here?
>
> Python 3.1.1 (r311:74480, Feb 3 2010, 13:36:47)
> [GCC 4.3.4] on linux2
>>>> -0.1 ** 0.1

> -0.7943282347242815
>>>> a = -0.1; b = 0.1
>>>> a ** b

> (0.7554510437117542+0.2454609236416552j)
>>>> -abs(a ** b)

> -0.7943282347242815
>
> Why does the literal version return the signed magnitude and the
> variable version return a complex?

Didn't we just do this one last week?

--
Grant Edwards grante Yow! Hello? Enema Bondage?
at I'm calling because I want
visi.com to be happy, I guess ...

Grant Edwards, Feb 11, 2010
6. ### AahzGuest

In article <hl1j4a\$j61\$>,
Grant Edwards <> wrote:
>
>Didn't we just do this one last week?

Let's do the Time Warp again!
--
Aahz () <*> http://www.pythoncraft.com/

"At Resolver we've found it useful to short-circuit any doubt and just
refer to comments in code as 'lies'. "

Aahz, Feb 12, 2010
7. ### Albert van der HorstGuest

In article <>,
Terrence Cole <> wrote:
>Can someone explain to me what python is doing here?
>
>Python 3.1.1 (r311:74480, Feb 3 2010, 13:36:47)
>[GCC 4.3.4] on linux2
>>>> -0.1 ** 0.1

Python 4.0
Warning: misleading blank space, expected:
- 0.1**0.1

>-0.7943282347242815
>>>> a = -0.1; b = 0.1
>>>> a ** b

>(0.7554510437117542+0.2454609236416552j)
>>>> -abs(a ** b)

>-0.7943282347242815
>
>Why does the literal version return the signed magnitude and the
>variable version return a complex?
>
>Cheers,
>Terrence
>

--
--
Albert van der Horst, UTRECHT,THE NETHERLANDS
Economic growth -- being exponential -- ultimately falters.
albert@spe&ar&c.xs4all.nl &=n http://home.hccnet.nl/a.w.m.van.der.horst

Albert van der Horst, Feb 22, 2010
8. ### Steven D'ApranoGuest

On Mon, 22 Feb 2010 18:01:44 +0000, Albert van der Horst wrote:

> In article <>,
> Terrence Cole <> wrote:
>>Can someone explain to me what python is doing here?
>>
>>Python 3.1.1 (r311:74480, Feb 3 2010, 13:36:47) [GCC 4.3.4] on linux2
>>>>> -0.1 ** 0.1

>
> Python 4.0
> Warning: misleading blank space, expected:
> - 0.1**0.1
>
>>-0.7943282347242815

Making spaces significant in that fashion is mind-bogglingly awful. Let's
look at a language that does this:

[steve@sylar ~]\$ cat ws-example.rb
def a(x=4)
x+2
end

b = 1
print (a + b), (a+b), (a+ b), (a +b), "\n"

[steve@sylar ~]\$ ruby ws-example.rb
7773

--
Steven

Steven D'Aprano, Feb 23, 2010
9. ### Mark DickinsonGuest

On Feb 23, 8:11 am, Steven D'Aprano
<> wrote:
> Making spaces significant in that fashion is mind-bogglingly awful. Let's
> look at a language that does this:
>
> [steve@sylar ~]\$ cat ws-example.rb
> def a(x=4)
>     x+2
> end
>
> b = 1
> print (a + b), (a+b), (a+ b), (a +b), "\n"
>
> [steve@sylar ~]\$ ruby ws-example.rb
> 7773

Hmm. That's pretty nasty, all right. Not that Python can claim to be
immune to such behaviour:

>>> 3 .real

3
>>> 3. real

File "<stdin>", line 1
3. real
^
SyntaxError: invalid syntax

Though the fact that one of the cases raises an exception (rather than
silently giving some different behaviour) ameliorates things a bit.

--
Mark

Mark Dickinson, Feb 23, 2010
10. ### Steven D'ApranoGuest

On Tue, 23 Feb 2010 05:48:09 -0800, Mark Dickinson wrote:

> On Feb 23, 8:11Â am, Steven D'Aprano
> <> wrote:
>> Making spaces significant in that fashion is mind-bogglingly awful.
>> Let's look at a language that does this:
>>
>> [steve@sylar ~]\$ cat ws-example.rb
>> def a(x=4)
>> Â  Â  x+2
>> end
>>
>> b = 1
>> print (a + b), (a+b), (a+ b), (a +b), "\n"
>>
>> [steve@sylar ~]\$ ruby ws-example.rb
>> 7773

>
> Hmm. That's pretty nasty, all right. Not that Python can claim to be
> immune to such behaviour:
>
>>>> 3 .real

> 3
>>>> 3. real

> File "<stdin>", line 1
> 3. real
> ^
> SyntaxError: invalid syntax
>
>
> Though the fact that one of the cases raises an exception (rather than
> silently giving some different behaviour) ameliorates things a bit.

It ameliorates it *completely* -- you won't get silent errors in Python
because you add or delete whitespace around a dot.

"I find it amusing when novice programmers believe their main job is
preventing programs from crashing. ... More experienced programmers
realize that correct code is great, code that crashes could use
improvement, but incorrect code that doesn't crash is a horrible
nightmare."

http://www.pphsg.org/cdsmith/types.html

The edge case occurs because dot does double-duty as an operator and as
part of float literals. However, float literals never include whitespace:

>>> 1.5

1.5
>>> 1 . 5

File "<stdin>", line 1
1 . 5
^
SyntaxError: invalid syntax

and likewise for 1. 5 and 1 .5 -- the only way to get a float literal
with a decimal point is by not including whitespace in it. So there is
never any ambiguity about floats. You can even do this:

>>> 1.5.__str__()

'1.5'

And since . is an operator outside of float literals, you can do this:

>>> import sys
>>> sys . platform

'linux2'

although why you'd want to escapes me

This actually is a feature, since it is useful when calling methods on
int literals. However this is a very rare thing to do.

--
Steven

Steven D'Aprano, Feb 24, 2010
11. ### AahzGuest

In article <>,
Steven D'Aprano <> wrote:
>
>[steve@sylar ~]\$ cat ws-example.rb

Ahhh, you're a Heroes fan. ;-)
--
Aahz () <*> http://www.pythoncraft.com/

"Many customs in this life persist because they ease friction and promote
productivity as a result of universal agreement, and whether they are
precisely the optimal choices is much less important." --Henry Spencer

Aahz, Feb 27, 2010