Posted by
eric ehlers on
URL: http://quantlib.414.s1.nabble.com/Java-wrapper-test-program-using-Eclipse-3-1-1-tp4555p4565.html
Hi Joe,
Many thanks for the additional info. Some things you might want to look at:
- Value Objects: This feature has been implemented in CVS for release
in OH version 0.1.4/QLA version 0.3.13. All QL/QLA objects stored in
OH have a corresponding VO which is generated automatically and stores
a snapshot of the input values to the object's constructor. The idea
is to extend OH to support serialization of VOs and this might tie in
with your requirement for the database schema to be able to hold the
data necessary to price derivatives. We'd need to consider the usual
questions which arise when interfacing an object oriented environment
such as QuantLib with a SQL database.
- On the subject of serialization please see also previous discussions
on that subject in the mailing list:
http://sourceforge.net/mailarchive/forum.php?thread_id=6507779&forum_id=4300http://sourceforge.net/mailarchive/forum.php?thread_id=6539684&forum_id=4300http://sourceforge.net/mailarchive/forum.php?thread_id=6540997&forum_id=4300http://sourceforge.net/mailarchive/forum.php?thread_id=6543552&forum_id=4300http://sourceforge.net/mailarchive/forum.php?thread_id=6700844&forum_id=4300- Source Generation: At present the source code for QLA functions is
generated automatically by python application srcgen based on metadata
stored in XML files. We should consider whether srcgen could be
extended to generate your UDFs.
If you'd like to formalize your current thoughts into a high-level
document I'd be happy to post that on the QLA website. It would be
good to publish the fact that this requirement has been identified,
for the benefit of anyone else who may come along with a similar
interest.
As to subsequent steps, at this point I don't anticipate personally
implementing this but I'd be happy to support any effort on the part
of yourself or anyone else who takes an interest.
Regards,
Eric
On 3/27/06, Joe Byers <
[hidden email]> wrote:
> Eric,
>
> I think fleshing it out a bit more is a good Idea. I think it requires a
> global objective for WHY this would be beneficial for an individual or firm.
> Also, an initial pass at a list of DB's for the functions is needed.
>
> I am in the rough start up phase of working on a proposal with two
> professors here at the Univeristy of Tulsa on a database schema for the
> financial industry that is Sarbanes Oxley compliant with audit trails. This
> schema will have a design for data warehousing of finanical security prices
> and derivative prices. I also want it to have the ability to hold the
> paramters for pricing the derivatives. That is the arguments quantlib
> functions needs. This way one can have a database of their positions or in
> prices for a research project. The default or modified arguments are saved
> and the portfolio can be revalued or research conducted easily on a few or
> several thousand derivative instruments.
>
> This ability to value a portfolio of derivative instruments inside a DB is
> what I see as the objective for this project. Individuals, researcher,
> and companies could benefit applicable to front, mid, and back office
> operations.
>
> Let me know where this needs to go from here.
>
> Have a good day.
>
> Joe