On Sun, Aug 10, 2014 at 8:58 PM, Stewart Wadsworth < stewart.wadswo...@gmail.com> wrote:
> Aaron, > > Thanks for the tip. _eval_evalf made it a lot simpler. > > The problem I'm having now is I'm trying to do some multiple inheritance > where I inherit from Symbol and another class that I've written. > > The issue I'm having is that Symbol overrides __new__ so (it feels like) > I'm basically forced into using the same signature for __new__ in my other > class which is not really desirable. > > I noticed that the only class that implements __new__ that Symbol inherits > from is Basic (and it inherits? from a few meta classs which implement > __init__ ... but that's getting pretty far over my head) I don't think I > can get away with not having a __new__ or __init__ in the other class I'm > trying to inherit from > Don't mind the metaclasses. A metaclass's __init__ gets called before the class's __new__. I'm not sure if you can reasonably change the __new__ constructor. Stackoverflow will probably provide more wisdom than I can. You'll need to make fortuitous use of super() most likely. Understanding how classes work in Python is a good start. I saw a good PyCon talk metaclasses that's a good introduction to both classes and metaclasses. I think https://www.youtube.com/watch?v=ANhTacigaf8 is the one. > > Could you point me in the right direction for getting around this? I'm > fine with always inheriting from sympy Basic if that would some how take > care of it. > That should work too. From a philosophical perspective, you shouldn't subclass from Symbol unless you actually want to use the same interface, which includes the argument to the constructor. > > Or do I need to implement __new_stage2__ in the other class or something? > I don't know what that does. I've never seen a subclass implement it. I think it has something to do with making assumptions work correctly. Ugh, I don't like this old assumptions metaclass nonsense in the core that no one can understand. Aaron Meurer > > > On Wednesday, July 30, 2014 8:05:32 PM UTC-4, Aaron Meurer wrote: > >> Don't monkeypatch Expr. If you're doing that, you're either doing >> something wrong, or there's something that needs to be added to SymPy. >> >> All you need to do is define _eval_evalf, and then calling evalf on an >> expression containing your subclass will call that method and use the >> result. I don't know if there is any documentation for it, but the >> signature is _eval_evalf(self, prec), where prec is the precision in >> decimal digits. >> >> If you want to replace it with an exact value, rather than a numerical >> value, you can define a method called doit() that returns that value. >> Then call doit() on the top-level expression. >> >> Aaron Meurer >> >> On Wed, Jul 30, 2014 at 10:24 AM, Stewart Wadsworth >> <stewart....@gmail.com> wrote: >> > Hello, >> > >> > I'm planning to use sympy for a project I'm working on and I have a few >> > questions... >> > >> > Basically I want to be able to associate a value with a symbol and then >> be >> > able to evaluate any function using that symbol for the value it's been >> > given. This would allow me to define properties in my classes as >> symbolic >> > functions but also be able to get the numeric value for the function at >> the >> > value specified for each symbol. >> > >> > Question 1: Is there already a built in way to do this without having >> to >> > call .subs() and .evalf() all over the place? (I'm pretty sure the >> answer is >> > "no" but I could have overlooked something....) >> > >> > Question 2: If not does this seem generally useful? Like something >> worth >> > incorporating into sympy? >> > >> > Question 3a: If it DOES seem like something worth incorporating: how >> can I >> > help? (I'm somewhat new to programming and don't have any formal >> > training...should I do a pull request and just do what I think is >> right?) >> > >> > Question 3b: If it DOES NOT seem like something worth incorporating: I >> > implemented this functionality by creating my own Symbol class and >> Monkey >> > patching the Expr class (copied the code below). Am I missing something >> > important here? I don't think this will break sympy... Is there a >> better way >> > to approach this? I'm a beginner so any feedback would be appreciated! >> > >> > >> > >> > CODE: >> > >> > class dSymbol(Symbol): >> > """ >> > Add a default value to Symbols >> > """ >> > >> > def __new__(cls, name, value, **assumptions): >> > return super().__new__(cls, name, **assumptions) >> > >> > def __init__(self, name, value, **assumptions): >> > self.default = value >> > >> > """ >> > Monkey patch the Expr class (inherited by Mul, Add,... all the function >> > classes in sympy i'm interested in using) >> > Add methods: >> > symvalue >> > value >> > """ >> > >> > def symvalue(self): >> > """ >> > returns the expression evaluated at the default value for each >> dSymbol >> > the result is still symbolic in that pi will be displayed as 'pi' >> not >> > '3.14159'... >> > """ >> > s = self.atoms(dSymbol) >> > for i in s: >> > self = self.subs(i, i.default) >> > return self >> > >> > def value(self, *args, **kwargs): >> > """ >> > returns the expression evaluated at the default value for each >> dSymbol >> > the result will be numeric (assuming it doesn't contain any >> 'Symbol') >> > """ >> > return self.symvalue().evalf(*args, **kwargs) >> > >> > Expr.symvalue = symvalue >> > Expr.value = value >> > >> > -- >> > 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 sympy+un...@googlegroups.com. >> > To post to this group, send email to sy...@googlegroups.com. >> > Visit this group at http://groups.google.com/group/sympy. >> > To view this discussion on the web visit >> > https://groups.google.com/d/msgid/sympy/b12b7e28-aff2- >> 4168-a3a2-c9f5ee67e1ad%40googlegroups.com. >> > For more options, visit https://groups.google.com/d/optout. >> > -- > 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 sympy+unsubscr...@googlegroups.com. > To post to this group, send email to sympy@googlegroups.com. > Visit this group at http://groups.google.com/group/sympy. > To view this discussion on the web visit > https://groups.google.com/d/msgid/sympy/8d82f942-979c-4a9a-ba06-72a41985583a%40googlegroups.com > <https://groups.google.com/d/msgid/sympy/8d82f942-979c-4a9a-ba06-72a41985583a%40googlegroups.com?utm_medium=email&utm_source=footer> > . > > For more options, visit https://groups.google.com/d/optout. > -- 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 sympy+unsubscr...@googlegroups.com. To post to this group, send email to sympy@googlegroups.com. Visit this group at http://groups.google.com/group/sympy. To view this discussion on the web visit https://groups.google.com/d/msgid/sympy/CAKgW%3D6JrYBnVVULQ481vSr9bfW8bsj8ZVkgRePv0QATQ0GvUQg%40mail.gmail.com. For more options, visit https://groups.google.com/d/optout.
