Hi Volker,
I put a new version of toric lattices patch. Now it should work nicely
with plain vectors and matrices, in particular homomorphisms seem to
do what they should (there is a simple example in the end of "module
level" documentation). I don't think anymore that there is any need
for a sepa
Hi Volker,
The first version (already documented and doctested) of toric lattices
is on trac:
http://trac.sagemath.org/sage_trac/ticket/9062
so you can take a look and comment if something is very bad. It is not
quite ready for review yet, but I don't anticipate changing existing
code - only addin
Hi Volker,
On May 21, 11:24 am, Volker Braun wrote:
> Ok, I agree that it would be nice for all derived cones to share the
> base cached data. How about Cone() just stores all of its cached data
> in a hash self.shared_cache (or so) instead of in self. For the
> computer thats essentially the sam
Ok, I agree that it would be nice for all derived cones to share the
base cached data. How about Cone() just stores all of its cached data
in a hash self.shared_cache (or so) instead of in self. For the
computer thats essentially the same as what you are suggesting, but in
the implementation you wo
On May 21, 5:24 am, Volker Braun wrote:
> Hi Andrey,
>
> Just to clarify, I only wanted to subclass Cones, but not Fans. The
> fan is a container class for cones and that doesn't change. In
> particular, I never felt the need to add any functionality at that
> level in my code.
>
> As for sharing
Hi Andrey,
Just to clarify, I only wanted to subclass Cones, but not Fans. The
fan is a container class for cones and that doesn't change. In
particular, I never felt the need to add any functionality at that
level in my code.
As for sharing of cached data, I think a clearer way would be to
defin
Hi Volker,
Fine, I am now convinced that you are right - we do need more
subclasses of cones ;-) And probably the same for fans?
Since I am still a bit paranoid about speed and memory consumption,
what do you think of the following scheme:
Make classes EnhancedFan and Enhanced Cone (not exposed
To summarize: Plain python cannot pickle inner classes or functions:
>>> class outer(object):
... class inner(object):
... pass
>>> o = outer()
>>> pickle.dumps(o)
(works)
>>> i = o.inner()
>>> pickle.dumps(i)
(ERROR)
The work-around is to make the cone factory a top-level class (or
functio
On 5/20/10 9:30 AM, Andrey Novoseltsev wrote:
You cannot pickle functions
According to
http://docs.python.org/library/pickle.html#what-can-be-pickled-and-unpickled
you can pickle certain types of functions. I don't know if that helps
at all, but your statement about pickling functions in
Hi Volker,
That definitely makes sense, let me think a bit about it and then post
the comments/agreements.
You cannot pickle functions, so you cannot pass a factory to construct
cones (because you may need to have more cones later, so you need to
store it). If you apply my cone patch and remove _
10 matches
Mail list logo
