Hi Ondrej,

I don't know much about solving recurrence relations on two variables, but
the sources
I am referencing provides the solutions for the recurrences involved with
Pell equation.
So, I think that would be enough.

Concerning the solutions of x**2 - 13*y**2 = 27, solutions returned by
diop_pell
are:
>>> diop_pell(13, 27)
[(220, 61), (40, 11), (768, 213), (12, 3)]

Here is how wolfram alpha represents it:
http://www.wolframalpha.com/input/?i=x**2+-+13*y^2+%3D+27

They provide general solutions for each class.
(Look in the box named "Integer solutions")

Regards,
Thilina




On Wed, Jul 3, 2013 at 2:55 AM, Ondřej Čertík <[email protected]>wrote:

> > I am really excited to tell you that I implemented the algorithm for
> > solving generalized Pell equation. For the past week or so I was working
> > on the case B**2 - 4*A*C > 0 in quadratic DEs. Now, since the pell
> > equation is solved, I can solve the above case by transforming it to a
> > Pell equation. I looked a bit at the transformation and it's not that
> hard.
> > I will be able to code it and finish implementing quadratic DEs at the
> > end of this week.
>
> Excellent, great job!
>
> >
> > Currently, solutions returned for the Pell equations are the basic
> solutions
> > of the particular equation passed to the Pell equation solver. We can
> > represent
> > other solutions by a recurrence. Both you and Aaron had answered on how
> > to represent the recurrence in the solution. If I am not mistaken
> rsolve()
> > currently
> > solves the recurrences in one variable. But recurrences we are talking
> here
> > involves two variables. So returning the recurrence itself won't be a
> good
> > idea.
> > What Wolfram alpha currently does is, it solves the recurrence and
> returns
> > the general solution without returning any other specific solutions.
> Would
> > that
> > be a bad idea since I am implementing lower level API's?
>
> Do you know how to solve the recurrence of two variables? Do you have
> some examples for Wolfram Alpha that you tried? Let me see some
> examples and think about the best way.
>
> > I coded the algorithms mostly looking at these two papers.
> >
> > [1] Solving the generalized Pell equation x**2 - D*y**2 = N, John P.
> > Robertson,
> > July 31, 2004, Pages 16 - 17 and 4 - 8. http://www.jpr2718.org/pell.pdf
> >
> > [2] Solving the equation ax**2 + bxy + cx**2 + dx + ey + f = 0, by John
> P.
> > Robertson. http://www.jpr2718.org/ax2p.pdf
> >
> > I added a commit. I would love to have your feedback on that. Please take
> > a look at that when you are free.
>
> I left some small comments.
>
> Ondrej
>
> >
> > Regards,
> > Thilina
> >
> >
> >
> >
> > --
> > You received this message because you are subscribed to the Google Groups
> > "sympy" 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/sympy.
> > For more options, visit https://groups.google.com/groups/opt_out.
> >
> >
>
> --
> You received this message because you are subscribed to the Google Groups
> "sympy" 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/sympy.
> For more options, visit https://groups.google.com/groups/opt_out.
>
>
>

-- 
You received this message because you are subscribed to the Google Groups 
"sympy" 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/sympy.
For more options, visit https://groups.google.com/groups/opt_out.


Reply via email to