On Mar 20, 2:41 pm, Ondrej Certik <ond...@certik.cz> wrote: > Yes, just search this list for examples of applications. Definitely > start on this as soon as possible and prepare a schedule etc. You can > then easily fit it on the template that PSF will put up. > > Ondrej
Well, Python posted their template, and there's not much to it (see link above). The other page says that I should create a page on the wiki. Does that still apply? So far, I have this list of things that I would like to do. I am still not sure how they will fit into a timeline: • Implement Issue 1336: I actually posted this. The idea is to create a arbitrary constant type that will aid with the simplification of ODE solutions. See the issue page (http://code.google.com/p/sympy/ issues/detail?id=1336) to see why this is important. I am not sure where this should fit in my timeline. On the one hand, it would make sense to implement this first so that I won't have to modify my code later on. On the other hand, it should be modular enough that this won't be an issue. But more importantly, I think that implementing it later on might be a better idea because this will require a stronger understanding of the sympy code base than the other items, which I should have more of later on in the summer. • Add the ability to solve for initial conditions: This should be straightforward. The solver engine is already there, I just have to modify the dsolve function to accept initial conditions arguments. This shouldn't take very long at all. • Add the ability to solve differential equations of the following forms/methods: - Separable - n-order linear homogeneous with constant coefficients. - Exact differential equation. - Some special cases of exact differential equations that aren't exact but can be made exact with a simple substitution. - Homogeneous differential equations (by this I mean equations that can be written in the form dy/dx=f(y/x) or dy/dx=f(x/y)). - n-Order non-homogeneous linear with constant coefficients via undetermined coefficients or variation of parameters: This should take longer than any other of the methods, as the solution is more algorithmic and less straight forward. - non-homogeneous linear with non-constant coefficients via reduction of order: perhaps this should go in the list below. • If I have time, add these lesser used types/methods: - Differential equations of the type dy/dx=p(x,y)/q(x,y) where p(x,y) and q(x,y) are linear functions, both of coincident and non-coincident lines. These are take straight from my ODE text. Am I missing anything important? I don't think I will have the time to do series , transformation, or numerical methods. • Clean up bugs (obviously): Should this fit into my timeline? I know that it will take up a significant amount of my time, but it doesn't seem like that would be continuous to me (e.g., "Weeks 5-6: fix bugs" would be kind of silly). • Format: In what order should the ODE engine try solution methods? Will some methods fail if they are tried before others are? Will some methods fail if the matcher catches them but they really can't be solved? I already have an example of this. The Bernoulli method I implemented, dsolve(f(x).diff(x)+x*f(x)-f(x),f(x)) catches f(x)^n with n=1, but the solution involves the power of 1/(1-n), so it returns the C1**(oo) in regular python and 0 ⎽⎽⎽⎽ ╲╱ C₁ in isympy (is this a bug or intentional?). I will submit a patch that checks to make sure that n≠1, but even then, it should be recognized as linear (f(x).diff(x)+x*f(x)-f(x)==f(x).diff(x)+(x-1)*f(x)). Also, speed and final format of the answer are considerations here. • Am I missing anything important here? --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "sympy" group. To post to this group, send email to sympy@googlegroups.com To unsubscribe from this group, send email to sympy+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/sympy?hl=en -~----------~----~----~----~------~----~------~--~---
