Re: [sage-devel] Re: Python vs Cython parents, and element constructors

2015-02-10 Thread Robert Bradshaw
On Tue, Feb 10, 2015 at 2:34 PM, Volker Braun wrote: > My advice would be to implement the parent in Python and benchnmark it. If > you are doing something in _element_constructor_ that would benefit greatly > by Cython then you can just move that part to Cython. +1 > On Tuesday, February 10, 20

[sage-devel] GSoC projects

2015-02-10 Thread William Stein
Does anybody have any GSoC projects to add to http://wiki.sagemath.org/GSoC/2015 There were basically zero that involved actual mathematics, so I've added two. One is relevant for quaternion algebras and another for modular forms. Hey people -- you could use this to support a frickin' grad

[sage-devel] Re: trac doesn't reset password

2015-02-10 Thread Volker Braun
Fixed by PM. On Tuesday, February 10, 2015 at 11:46:38 PM UTC+1, kates...@gmail.com wrote: > > Thank you! However... I received an email (as when I use the forgot > password link) but the new password doesn't work. > > Kate. > > On Tuesday, February 10, 2015 at 3:32:44 PM UTC-7, Volker Braun wr

[sage-devel] Re: trac doesn't reset password

2015-02-10 Thread katestange
Thank you! However... I received an email (as when I use the forgot password link) but the new password doesn't work. Kate. On Tuesday, February 10, 2015 at 3:32:44 PM UTC-7, Volker Braun wrote: > > I tried the new "reset passwords" button in the trac admin panel, let me > know if it worked ;-

[sage-devel] Re: Python vs Cython parents, and element constructors

2015-02-10 Thread Volker Braun
My advice would be to implement the parent in Python and benchnmark it. If you are doing something in _element_constructor_ that would benefit greatly by Cython then you can just move that part to Cython. On Tuesday, February 10, 2015 at 8:20:08 PM UTC+1, Marc Mezzarobba wrote: > > Hi, > > W

[sage-devel] Re: trac doesn't reset password

2015-02-10 Thread Volker Braun
I tried the new "reset passwords" button in the trac admin panel, let me know if it worked ;-) On Tuesday, February 10, 2015 at 10:20:48 PM UTC+1, kates...@gmail.com wrote: > > I'm having the same problem! I'm katestange. Could you help me out? > > On Thursday, August 7, 2014 at 4:44:16 PM U

Re: [sage-devel] Re: self in the INPUT section of the docstring

2015-02-10 Thread Volker Braun
I tend to add it to OUTPUT, like OUTPUT: Returns the frob of x and y. If ``self`` is not in the range from 3 to 7 then a ``ValueError`` is raised. On Tuesday, February 10, 2015 at 10:41:49 AM UTC+1, Jeroen Demeyer wrote: > > On 2015-02-10 02:32, Volker Braun wrote: > > I read that but the nu

[sage-devel] Re: trac doesn't reset password

2015-02-10 Thread katestange
I'm having the same problem! I'm katestange. Could you help me out? On Thursday, August 7, 2014 at 4:44:16 PM UTC-6, Volker Braun wrote: > > done > > On Thursday, August 7, 2014 10:22:26 PM UTC+1, Vincent Knight wrote: >> >> Would it be possible to reset my password also? I've had the same probl

[sage-devel] Python vs Cython parents, and element constructors

2015-02-10 Thread Marc Mezzarobba
Hi, What is the recommended way to implement a parent that needs a fast element constructor but otherwise can be expected not to be queried too often once the coercion maps involving it are set up ? (I would like MyParent(obj) to have as little overhead as possible, not just coercions.) I can

[sage-devel] Re: self in the INPUT section of the docstring

2015-02-10 Thread Simon King
On 2015-02-10, Nathann Cohen wrote: >> In this case, the question remains: how should we document any special >> conditions on "self" if not in the INPUT block? > > In graphs we sometimes add notes or warnings, like: > > .. WARNING:: > > The graph must be connected Sometimes I used an ASSUMPT

Re: [sage-devel] Re: self in the INPUT section of the docstring

2015-02-10 Thread Nathann Cohen
> In this case, the question remains: how should we document any special > conditions on "self" if not in the INPUT block? In graphs we sometimes add notes or warnings, like: .. WARNING:: The graph must be connected For user-level functions we perform a check, si that an exception is raised

Re: [sage-devel] Re: self in the INPUT section of the docstring

2015-02-10 Thread Jeroen Demeyer
On 2015-02-10 02:32, Volker Braun wrote: I read that but the number of occurences is so tiny compared to the size of Sage that I would count that as an argument against. Also, e.g. the Google and NumPy Python docstring style explicitly says that you shall not In this case, the question remains