[sympy] ManagedProperties strangeness

2015-03-10 Thread Joachim Durchholz
ManagedProperties is declared like this: class ManagedProperties(with_metaclass(BasicMeta, BasicMeta)): ... I.e. ManagedProperties has BasicMeta both as its metaclass and as its superclass (so it is a metaclass itself). What's the purpose of this construction? I suspect we should

Re: [sympy] GSoC Group Theory

2015-03-10 Thread Jennifer White
Thanks for your response. The two areas that I had particular interest in were representing groups symbolically as you mentioned, and also potentially looking into generating Galois Groups. Do you think these would be good areas to start with? Obviously there is a huge amount that could be done

Re: [sympy] Re: Slow inverse_laplace_transformation

2015-03-10 Thread Aaron Meurer
So perhaps you would have better results if you started with a Float with higher precision, like Float(0.07071, 100). Aaron Meurer On Sun, Feb 8, 2015 at 2:57 PM, Kalevi Suominen jks...@gmail.com wrote: On Sunday, February 8, 2015 at 12:06:46 AM UTC+2, Sten Sogaard wrote H =

[sympy] GSoC - Make Sage use CSymPy as the symbolic engine

2015-03-10 Thread Isuru Fernando
Hi, I'm Isuru Fernando. I'm a 3rd year Computer Science and Engineering undergraduate at University of Moratuwa. I've been working on CSymPy for some time now and would like to work on the Sage GSoC project to use CSymPy as the symbolic engine. I have a couple of questions regarding the

[sympy] Re: What am I doing wrong with apart()?

2015-03-10 Thread Francesco Biscani
A much reduced testcase: x,y = [Symbol(_) for _ in ['x','y']] foo = 1/(x+y)-Rational(1,2)*x**2 print(foo.apart(x)) This prints: 1/(x + y) If the Rational has an integral value, everything looks ok: foo = 1/(x+y)-Rational(6)*x**2 print(foo.apart(x)) Output: -6*x**2 + 1/(x + y) On 10 March

Re: [sympy] What am I doing wrong with apart()?

2015-03-10 Thread Aaron Meurer
It looks like a bug. If you do foo.expand() - foo.apart(xi_s).expand() you get a term which is definitely not equal to zero. Aaron Meurer On Tue, Mar 10, 2015 at 9:57 AM, Francesco Biscani bluesca...@gmail.com wrote: Hello list, I feel a bit embarrassed as probably I am doing something

[sympy] What am I doing wrong with apart()?

2015-03-10 Thread Francesco Biscani
Hello list, I feel a bit embarrassed as probably I am doing something blatantly wrong, but I cannot understand what is going here. Consider this snippet (copy pasted from an Ipython notebook): import sympy sympy.init_printing() from sympy import Symbol,sqrt,cos,sin,solve,Function,atan,apart,S

Re: [sympy] GSoC Group Theory

2015-03-10 Thread Aaron Meurer
Those both sound like good ideas. I would look into what the prerequisites for each are so you can see what is already implemented and what needs to be implemented. Aaron Meurer On Tue, Mar 10, 2015 at 8:27 AM, Jennifer White jcrw...@googlemail.com wrote: Thanks for your response. The two

Re: [sympy] Is BasicType useful?

2015-03-10 Thread Aaron Meurer
On Tue, Mar 10, 2015 at 3:02 PM, Joachim Durchholz j...@durchholz.org wrote: Am 10.03.2015 um 20:25 schrieb Aaron Meurer: I think this explains the issue http://docs.sympy.org/latest/guide.html#functions (search for meta). That's outdated, MetaBasicMeths does not exist anymore (it might

Re: [sympy] Is BasicType useful?

2015-03-10 Thread Joachim Durchholz
Am 10.03.2015 um 21:10 schrieb Aaron Meurer: By the way, I still think we should remove this document https://github.com/sympy/sympy/pull/2807. Heh. I agree with rather no docs than bad docs. I have no idea who could work on salvaging what's still useful. But maybe if there is nobody around

Re: [sympy] Is BasicType useful?

2015-03-10 Thread Aaron Meurer
I think this explains the issue http://docs.sympy.org/latest/guide.html#functions (search for meta). By the way, I still think we should remove this document https://github.com/sympy/sympy/pull/2807. Aaron Meurer On Tue, Mar 10, 2015 at 2:16 PM, Joachim Durchholz j...@durchholz.org wrote: Am

Re: [sympy] ManagedProperties strangeness

2015-03-10 Thread Aaron Meurer
I believe the purpose of this class was to make it easier to remove the assumptions from the core. I would dig into the pull request and commits that introduced it to see if there is more information. I agree it's weird to have BasicMeta be both the metaclass and the superclass. Aaron Meurer On

Re: [sympy] Is BasicType useful?

2015-03-10 Thread Aaron Meurer
I suppose it must be there to distinguish subtypes from other subtypes of BasicMeta that aren't BasicType. Either that or it used to have methods but they were deleted, and the class was not deleted. I would look into the git history of it using git blame. Aaron Meurer On Tue, Mar 10, 2015 at

Re: [sympy] piecewise_fold of nested piecewise expressions; bug?

2015-03-10 Thread claytongdavis
OK Aaron -- thanks for your reply. I'll see if I can generate a patch... On Monday, March 9, 2015 at 8:03:11 PM UTC-4, Aaron Meurer wrote: It's probably an unimplemented feature. There are a lot of things that piecewise_fold can't handle which it could. Aaron Meurer On Mon, Mar 9, 2015

[sympy] Is BasicType useful?

2015-03-10 Thread Joachim Durchholz
Class BasicType defines no methods or attributes, its only subclass is BasicMeta. Neither class is in the sympy namespace. Does BasicType serve any purpose that BasicMeta would not? (Context: I found this while exploring SymPy's metaclass hierarchy. Metaclass status should not affect the

Re: [sympy] ManagedProperties strangeness

2015-03-10 Thread Ronan Lamy
Le 10/03/15 09:46, Joachim Durchholz a écrit : ManagedProperties is declared like this: class ManagedProperties(with_metaclass(BasicMeta, BasicMeta)): ... I.e. ManagedProperties has BasicMeta both as its metaclass and as its superclass (so it is a metaclass itself). What's the purpose of