kcrisman wrote :
> On Sep 16, 4:04 pm, Francois Maltey <[email protected]> wrote:
>   
>> I play with sage, exp, sin, cos, sinh, and co...
>>
>> var("a,b,c")
>> exp(a)^2 # returns exp(2a) is right
>> exp(a)^(1/2) # returns exp (a/2) is wrong, with a=2*i*pi we get -1=1
>> exp(a)^b # returns exp(a*b) is wrong
>>     
>
> Well, there is a unique exp, but not a unique square root (or in
> general other power, since they may be defined using log).
But is XXX^(1/2) are unique and not multivariate : sage remains 
(a^2)^(1/2) and (a^b)^c.
Neither automatic convert (a^2)^(1/2) to a nor (a^b)^c to a^(b*c)

I used both Axiom with only multivariate functions asin sin a = a and 
(a^2)^(1/2)=a
and Maple/mupad with no multivariate functions (but bugs).

Mathematics for undergraduate are finest by this way.
 
>  Are you suggesting that exp(a)^(1/2) always return exp(a)^(1/2)
It's what I prefer
> , or that it return something about branches?
In my mind log z = ln |z| + i arctan2 (Re(z),Im(z)) where arctan2 (x,y) 
in ]-pi,pi].
This logarithm is MONO-variate, but we can't write ln (u v) = ln u + ln 
v in complex domain.

> sage: (-1)^(1/3)
> (-1)^(1/3)
>   
It's right, arctan(-1,0)=pi, e^(i*pi/3) = 0.5 + 0.866... I
> But in general Sage does things over complex numbers fairly
> consistently.  We constantly get complaints about
> sage: (-1.)^(1/3)
> 0.500000000000000 + 0.866025403784439*I
>   
All right, so (exp (a))^b is a singular exception... And I prefer a 
system without such exception.

F.


--~--~---------~--~----~------------~-------~--~----~
To post to this group, send email to [email protected]
To unsubscribe from this group, send email to 
[email protected]
For more options, visit this group at 
http://groups.google.com/group/sage-support
URLs: http://www.sagemath.org
-~----------~----~----~----~------~----~------~--~---

Reply via email to