Re: does anyone use java+swig bindings?

Posted by Richard Gomes on
URL: http://quantlib.414.s1.nabble.com/does-anyone-use-java-swig-bindings-tp1326p1335.html



Tito Ingargiola wrote:

>
> Hi Richard,
>
> I am hoping to hear that someone is able to use the swig bindings for
> something non-trivial in java which is clearly more than you or I have
> (yet!) managed.  I want to see some code that actually works as I'm
> skeptical that the bindings are functional for java.
>
> Luigi's response to you seems correct for the case where one wants to
> access functionality which hasn't yet been mapped in the swig bindings.  
> My issue is instead one where the swig bindings are defined but they
> just don't seem to work.  I agree that writing my own JNI stuff wouldn't
> be fun, but I wouldn't try to map the entire api, just specific
> questions I might want answered, e.g., the price of an option on a
> future.  This is a much more constrained issue which isn't so difficult
> much as I'd prefer to avoid it if the swig stuff can be made to work...
>
> All of that said, if you have a concrete idea about how we can work
> together, I'm glad to hear it.  For now, it seems a starting point is to
> get something simple, like what I'd provided, to work.  I hope someone
> will be able to point us in the direction of a working example.  (Pretty
> please! :-)
>
> Regards,
>
>     Tito.
>
> ps - I've moved this from the dev to the users list as it seems the more
> appropriate forum
>
> */Richard Gomes <[hidden email]>/* wrote:
>
>     Hi Tito,
>
>     I have a lot of code in Java as well and I'd like to plug QuantLib in.
>
>     It's incredible how you and me are facing the same problems, following
>     the same path and started translating the same example! :D When I
>     looked at you Java code it was quite similar to mine (of course... you
>     also translated the example in C++).
>
>     Oh well...
>
>     I dont think that writing your own JNI is a good idea.
>     It's a huge, boring and error prone task. I'd rely on SWIG instead.
>     It'd be much easier to learn how to hack the intrincacies of SWIG.
>
>     It would be great if we were able to convert comments from C++ code to
>     JavaDOC as well, as Eclipse would be able to show documentation as
>     you type.
>
>     Another thing to be considered is performance.
>     I tried to use Matrix and it does not make sense to initialize every
>     cell at a time. The performance will be poor. It means that we should
>     create additional methods intended to be called by another languages
>     due
>     to poor performance of separated calls.
>
>     I've posted one thread quite similiar to yours and Luigi gave me some
>     directions. In case, I've use Matrix as an example of issues I've found.
>
>     I'm planning to work on SWIG this weekend. Maybe we could work together
>     on this kind of thing.
>
>     Kind Regards
>
>     -- Richard Gomes
>
>
>     -------------------------------------------------------------------------
>     This SF.net email is sponsored by: Splunk Inc.
>     Still grepping through log files to find problems? Stop.
>     Now Search log events and configuration files using AJAX and a browser.
>     Download your FREE copy of Splunk now >> http://get.splunk.com/
>     _______________________________________________
>     QuantLib-dev mailing list
>     [hidden email]
>     https://lists.sourceforge.net/lists/listinfo/quantlib-dev
>
>

Tito,

I'm still in the beginning but I managed to calculate valuation of
options using Black-Scholes (from the sample translated into Java).
Unfortunately I couldnt make it work for all engines. I'm still trying
to solve but didnt have enough time to hack why Quantlib (C++ code) is
throwing a certain exception.

My idea is focus on something I need to make Java work fine, not
review/validate/implement everything. This is not my aim.

About working together.
 From your previous email, I'm sure we are both very focused, busy and
independent. I think the best alternative is simply share experiences
and send our patches to Luigi.

Regards

--
Richard Gomes
http://www.linkedin.com/in/rgomes
M: +44(77)9955-6813
H: +44(870)068-8205
W: +44(20)7378-3924


-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >>  http://get.splunk.com/
_______________________________________________
QuantLib-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/quantlib-users