Re: Google Summer Of Code, TISEAN package

[Dag W]

Hello Pablo,

Thanks for the answer.


Juan Pablo Carbajal-2 wrote
> It is not an intensive work.

I want to have something to do in my free time and improve my programming
skills. I was thinking of doing a small project first to familiarize myself
with Octave.



Juan Pablo Carbajal-2 wrote
> PS: also note that as a community we are interested on developers that
> might stick around after the complete their GSoC project. Would you
> like to stay as a maintainer of this package and help with the
> development of Octave? 

If it works out, I might start a more difficult project next. In case that
it is really interesting, I will definitely be around in the future for a
long time to make this a win-win scenario.



Juan Pablo Carbajal-2 wrote
> Also notice that your statistician skill wont be very useful at this
> point , because we are not developing the package further, we are just
> updating its interface to make it usable in Octave.

I have masters in statistics and time-series analysis (TSA) is one of my
specialties. I studied a course in nonlinear time-series analysis 10 years
ago, so I want to see if there is more to discover in this subject.



Juan Pablo Carbajal-2 wrote
> the package could
> be further developed,of course, but the first step is to bring it back
> from the crypt. If you haven't used the package before or read the
> book, you might find it very interesting (not required though).

The purpose, as I understand it, is to write a new Octave package for
nonlinear TSA. I was thinking of beginning by compiling Octave source and
TISEAN source and then start from there.

First, I am trying to find out how to write an Octave package. There is
information on the  contents of Octave packages
<https://www.gnu.org/software/octave/doc/interpreter/Creating-Packages.html#Creating-Packages>
 
. There is no information on which Octave source files to import, and how to
compile and link though. I have downloaded the source code of an existing
package to reverse-engineer it somehow.



Juan Pablo Carbajal-2 wrote
> The interface of these functions is adapted to a file based input and
> has a lot of "Fortranity" on it, so the signature of some function
> will need some redesign.
> It is not an intensive work. 

I prefer C and C++. I will replace Fortran code using the Octave C/C++ API
if possible or external C++ libraries if necessary. Is there any merit in
using old legacy code?



--
View this message in context: 
http://octave.1599824.n4.nabble.com/Re-Google-Summer-Of-Code-TISEAN-package-tp4668056p4668111.html
Sent from the Octave - Maintainers mailing list archive at Nabble.com.