Re: [Gretl-devel] Several things
> Google summer of code: one of the people I met in person for the
> first time at the conference was Stefano Iacus, from the R core
> team, who is one of the maintainers of the foreign package. In
> the past, we grabbed code from one another: they got the Eviews
> filter from Allin's implementation, we got the Stata filter,
> IIRC we both got SPSS from PSPP. We agreed that it'd be great to
> have a separate project for filters as a C library that gretl
> and R could both link to centralise efforts; however, it'd be
> quite a lot of work. A nice idea that Stefano had was to propose
> this as a GSOC project, with joint R/gretl sponsorship; however,
> I understand that there's little time left to set up this kind
> of initiative, so if you guys think it's a good idea, it's time
> to start planning it.
Sounds great. Personally I don't expect to have time to plan
this, but if anyone else is willing to take the initiative on
gretl's behalf I'll be glad to "sign on".
I am glad that Jack
brought up this issue. Last year, I have mentioned
personally to Allin that applying to GSoC could be a good idea. It was
too late to do it so, I had offered that I could help organizing this
next (this) year. I haven't done this before but I think what needs to
be done is we have to determine several potential projects that an
interested CS student can undertake over the summer. Then, we need to
apply to Google with a well written info on Gretl (and R if this is to
be a joint effort) along with the proposed projects.
So, first we need to discuss what would be some of the useful projects
that we can propose. I think having the seperate library for filters
for SPSS, Stata etc. sounds great for a project. So that is one. I can
think of some others. What do you think of:
1)- Importing more statistical distributions to gretl using R's
stats-library. Gnumeric is using R's library and that way they have
the most accurate functions regarding statistical distributions and
their inverses among the various spreadsheet software. I know this
from my own published study:
http://ideas.repec.org/a/eee/csdana/v52y2008i10p4579-4586.html
I think we can have lots of new statistical distributions and their
inverse functions if a student imported some of the code or
implemented the algorithms from R or some other resource.
2)- I remember that at one point Allin asked for help regarding
setting up a nicer and more detailed preferences menu where many
things that are adjustable only with the command line is accessible
from the GUI. I think a project involving this together with maybe
some other GUI improvements can be a good idea.
3)- Another one is improving the gnuplot integration so that more and
better looking graphs can be produced. I think visualization is very
important for a program like gretl and I don't think we are using
gnuplot to its potential (unless we learn gnuplot's own language). R
is great in easily producing a large selection of beautiful looking
plots and so should gretl, using its brilliant graphical interface.
4)- How about implementing or improving some simple statistical
functionality? Simple things are useful especially in a classroom
environment but their implementation can be delayed or neglected
because the resources are limited. How about improving ANOVA for
example?
5)- Related to the above, how about implementing the "confidence band
plots for mean Y and individual Y values" as in Gujarati 4th edition
p.144? I really like the confidence elipse and this one is also very
nice. I find this especially useful in teaching and I think it would
be good to have it.
I am sure there are other things that you can think of. Once we decide
on a list of potential projects, I can write a detailed description
for each of them. Then, maybe Allin or Jack can help me refine them,
which I can submit to Google afterwards. Once we apply to GSoC, can
handle the various correspondence and bureacracy work. We also need to
assign a mentor for each project. I can volunteer for mentoring too if
it is something that I can do.
What do you think?
Cheers
Talha
--
"Remember not only to say the right thing in the right place, but far
more difficult still, to leave unsaid the wrong thing at the tempting
moment." - Benjamin Franklin (1706-1790)
--