On Feb 20, 4:10 am, tomanizer <thomas.haede...@mailwitch.com> wrote:
Hi,
> If I remember correctly it took around 14 hours, but my memory might
> not be reliable.
Ouch.
> I will try to build it for the new Sage version and let you know how
> long it takes.
Yeah, it needs way too much RAM to build on smaller systems, i.e. 2GB
with gcc 4.1.2. And we have people regularly complain about LinBox
which uses about 350 to 700 MB at peak to compile.
> What do we need to do to make QuantLib a supported package?
Well, it kind of is, but it is considered experimental, i.e. back at
Sage Days 9 when Phaedon and I worked on them we ran into trouble with
boost, OSX and the non-system Python. So it might be worthwhile to
update it so it works on OSX, but it isn't a priority. If you want
some pointers I am happy to give them to you. But making it a standard
spkg in Sage won't happen for various reasons, the biggest one is just
the sheer size and the resources it needs to be build.
When I first played with Quantlib is was very impressed by its website
and seemingly large number of users, but the first time I compiled it
I instantly knew that Quantlib and Sage won't share a future together.
It is a prime example why boost and templates are a bad idea when used
in this way since the compile time is completely unacceptable.
> I am interested in helping, but I am not a Python/Sage genius (yet).
Excellent.
> Generally, is there some interest in taking Sage further as a platform
> for Quantitative Finance?
Absolutely. There is already a Google group called sage-finance, but
it has been quiet there for a while. Note that Sage contains excellent
support for generalized hidden markov models for example. Phaedon is
working on options pricing and also date functions and he promised to
send some code soon :)
> I remember William mentioning in a post some time that he would like
> to see more finance support in Sage too.
>
> If more people are interested we could
> - support QuantLib on Sage and extend the SWIG interface (e.g.
> docstrings, proper types, etc)
The SWIG interface is autogenerated code, so this will be tough. I
think given the above situation that the next suggestion is a much
better use of your time.
> - port QuantLib to Python/Sage
That is partially happening, but I am not sure what people want.
> - create a new Quant Finance Platform for Sage
Yes, all the way :)
> Any suggestions? Any more volunteers?
Glenn is working in this area, too, but I am not sure if he is reading
this thread. We have a CUDA box up and running and will be working on
getting code for las vegas models onto the GPU to get nice speedups :)
Cheers,
Michael
--~--~---------~--~----~------------~-------~--~----~
To post to this group, send email to sage-support@googlegroups.com
To unsubscribe from this group, send email to
sage-support-unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/sage-support
URLs: http://www.sagemath.org
-~----------~----~----~----~------~----~------~--~---
