I want to express an equality like x=c as inequalities  x -c <= 0 , x -c >= 
0  .   Having x = oo  is a valid constraint, which is equiv to x -oo <= 0 
and x =oo >= 0.  Sure I can get rid of x -oo <= 0  but the whole point is 
that SAGE converts  x-oo<=0 and x = oo >=0  to x - oo <= 0 and -Infinity >= 
0.  Which clearly kills the original semantics of (x -oo <= 0 and x =oo >= 
0).  

Also,  why not have such a constraint ?   it could mean I don't know the 
bound of x so I will totally over approximate it (to something that's 
always valid)  

On Tuesday, March 26, 2013 2:02:23 AM UTC-6, P Purkayastha wrote:
>
> On 03/26/2013 12:59 PM, tvn wrote: 
> > In my project I often have inequalities x <= c , x+y <= c, x -y <= c and 
> > to make things uniform, I move all the consts to the lhs, e.g. x <= c 
> > becomes x - c <= 0  .  However this becomes problem when c is Infinity 
> > because Sage automatically converts x - oo <= 0  to -Infinity <= 0  and 
> > removes all variables in the original relation.  Is there a way to get 
> > around this ? I thought about creating a new variable to represent oo 
> > but thought there could be easier way. 
> > 
> > In fact I am not so sure if such action is desirable,  for example if I 
> > have   x <= oo and x >= oo  ,  this clearly implies x = oo .  But treat 
> > these as -oo >= 0 and -oo <= 0  make no sense (-oo >= 0 is False). 
> > Sympy does not use this action and allows relations such as x - oo <= 0 
>  . 
>
> Mathematically speaking, why would you want to have a constraint saying 
> x <= oo? This should behave the same as not having any constraint on x. 
> Perhaps I am not understanding what you want to constrain here. 
>
>
>
>

-- 
You received this message because you are subscribed to the Google Groups 
"sage-support" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To post to this group, send email to [email protected].
Visit this group at http://groups.google.com/group/sage-support?hl=en.
For more options, visit https://groups.google.com/groups/opt_out.


Reply via email to