GSOC
Joshua Herman
---- LOOK ITS A SIGNATURE CLICK IF YOU DARE---
http://www.google.com/profiles/zitterbewegung
Burcin Erocal
On Tue, 23 Feb 2010 23:41:12 -0600
Joshua Herman <zitterb...@gmail.com> wrote:
> Is sage doing GSOC?
Yes! But we should start organizing early this year.
Last year we started the planning too late and didn't manage to arrange
a proper list of tasks and mentors in time for the application. Things
have improved in the meanwhile since there are already a few task lists
(e.g., notebook [1], symbolics[2], polyhedra[3]) on the wiki. I'm sure
we can make a wiki page pointing to all these and recycle the forms
from last year for an application. Does anybody know the deadlines?
[1] http://wiki.sagemath.org/devel/SageTasks
[2] http://wiki.sagemath.org/symbolics
[3] http://wiki.sagemath.org/PolyhedraWishList
> I was thinking I could make a mobile edition for
> sage.
What do you mean by a mobile edition? Do you want to make the notebook
compatible with different browsers and touchscreen interfaces or are
you interested in making Sage run on exotic devices?
Any of these would be a great contribution to Sage. Thanks for your
interest.
Cheers,
Burcin
Harald Schilly
> Yes! But we should start organizing early this year.
And since sage wasn't accepted 3 times in a row (afaik) we should be
careful what we request. I think we have to drop all math-related
tasks. Just focus on things like "management", "server-
infrastructure", "webserver-design-blahh", ... and clear descriptions
with a good explanation what the result should be. Also a profile what
an interested student needs to know about. I'm willing to help writing
some text, ...
h
Joshua Herman
1. Make a mobile version of the notebook interface (Able to be used by
a mobile browser)
2. Get it to compile on arm and work on an android (If I have time)
---- LOOK ITS A SIGNATURE CLICK IF YOU DARE---
http://www.google.com/profiles/zitterbewegung
> --
> To post to this group, send an email to sage-...@googlegroups.com
> To unsubscribe from this group, send an email to sage-devel+...@googlegroups.com
> For more options, visit this group at http://groups.google.com/group/sage-devel
> URL: http://www.sagemath.org
>
Joshua Herman
would need help with writing the application document
https://docs.google.com/Doc?docid=0AWHfMu95Txe_ZHRrNmNjY181MDFkZzVxNXBkMg&hl=en
---- LOOK ITS A SIGNATURE CLICK IF YOU DARE---
http://www.google.com/profiles/zitterbewegung
Sent from Chicago, Illinois, United States
Peter Jeremy
>On Feb 24, 12:01 pm, Burcin Erocal <bur...@erocal.org> wrote:
>> Yes! But we should start organizing early this year.
>
>And since sage wasn't accepted 3 times in a row (afaik) we should be
>careful what we request. I think we have to drop all math-related
>tasks. Just focus on things like "management", "server-
>infrastructure", "webserver-design-blahh", ... and clear descriptions
>with a good explanation what the result should be.
Probably we need to have a closer read of where GSOC funding is
targetted. There may be a way to creatively word the proposals to
enhance their chances of being funded.
My other suggestion for GSOC tasks would be the various the porting
efforts (Windows, Solaris, FreeBSD). These don't require in-depth
maths skills and so could be attractive to a wider range of students.
--
Peter Jeremy
William Stein
Yes, I think there should be a Sage GSOC application. I'm OK with
being listed as a mentor. However, I won't write the actual
application, since I wrote three applications in a row that were all
turned down, so clearly I don't know how to write one.
What happened last year, by the way, was that several people wanted to
write a GSOC application. However, nobody volunteered to mentor any
Sage-related GSOC projects at all. As a result, we did not even apply
to have Sage as a mentoring organization.
-- William
Nils Bruin
> What happened last year, by the way, was that several people wanted to
> write a GSOC application. However, nobody volunteered to mentor any
> Sage-related GSOC projects at all. As a result, we did not even apply
> to have Sage as a mentoring organization.
I will be supervising a 3rd year mathematics/computer science student
this summer to work on things that might end up in sage, so I am very
much interested in projects with as many of the following
characteristics as possible:
1) Allows the student to learn some interesting mathematics
2) does not require too much mathematics background
3) does not require knowledge of too many different components (he'll
only have 4 months, so if he needs all of them to familiarize himself
with the code then he won't be able to do anything interesting)
4) high potential for rewarding results
The lists Burcin pointed to above are already very useful, but if
people know other lists or problems, please let me know.
Joshua Herman
Are you guys interested in adapting phcpack for sage? Its a solver for polynominal homotopy?
On Feb 24, 2010 5:51 PM, "Nils Bruin" <nbr...@sfu.ca> wrote:
On Feb 24, 2:48 pm, William Stein <wst...@gmail.com> wrote:
> What happened last year, by the way, w...
I will be supervising a 3rd year mathematics/computer science student
this summer to work on things that might end up in sage, so I am very
much interested in projects with as many of the following
characteristics as possible:
1) Allows the student to learn some interesting mathematics
2) does not require too much mathematics background
3) does not require knowledge of too many different components (he'll
only have 4 months, so if he needs all of them to familiarize himself
with the code then he won't be able to do anything interesting)
4) high potential for rewarding results
The lists Burcin pointed to above are already very useful, but if
people know other lists or problems, please let me know.
--
To post to this group, send an email to sage-...@googlegroups.com
To unsubscribe from this gr...
Joshua Herman
Link to http://www.math.uic.edu/~jan/download.html
On Feb 24, 2010 11:12 PM, "Joshua Herman" <zitterb...@gmail.com> wrote:
Are you guys interested in adapting phcpack for sage? Its a solver for polynominal homotopy?
>
> On Feb 24, 2010 5:51 PM, "Nils Bruin" <nbr...@sfu.ca> wrote:
>
> On Feb 24, 2:48 pm, William Stein <wst...@gmail.com> wrote:
> What happened last year, by the way, w...
>
> I will be supervising a 3rd year mathematics/computer science student
> this summer to work on ...
Mike Hansen
> Link to http://www.math.uic.edu/~jan/download.html
>
> On Feb 24, 2010 11:12 PM, "Joshua Herman" <zitterb...@gmail.com> wrote:
>
> Are you guys interested in adapting phcpack for sage? Its a solver for
> polynominal homotopy?
phcpack is already an optional package, and there is an interface to
it in sage/interfaces/phc.py, but maybe Marshall Hampton has a better
idea of improvements / work that could be done with it.
--MIke
Joshua Herman
---- LOOK ITS A SIGNATURE CLICK IF YOU DARE---
http://www.google.com/profiles/zitterbewegung
> --
> To post to this group, send an email to sage-...@googlegroups.com
David Joyner
One thought is to make it easier to use the graph theory functionality to do
things useful in social network analysis. This would involve R and Sage and
NetworkX (all standard components). Lots of people are interested in
SNA, from google to the military. It would require some knowledge of
probability, statistics, and graph theory.
Nathann Cohen
http://jung.sourceforge.net/doc/api/index.html
They are particularly fond of measures of clustering and metrics that
I am not too impatient to rewrite in Sage, though it is not
necessarily a tough job :-)
Nathann
On Feb 25, 1:40 pm, Minh Nguyen <nguyenmi...@gmail.com> wrote:
> Hi David,
>
> On Thu, Feb 25, 2010 at 11:07 PM, David Joyner <wdjoy...@gmail.com> wrote:
>
> <SNIP>
>
> > One thought is to make it easier to use the graph theory functionality to do
> > things useful in social network analysis. This would involve R and Sage and
> > NetworkX (all standard components). Lots of people are interested in
> > SNA, from google to the military. It would require some knowledge of
> > probability, statistics, and graph theory.
>
> NetworkX provides many useful features for social network analysis
> (SNA). A pressing concern at the moment is to get the NetworkX spkg in
> Sage upgraded to at least version >= 1.0; see ticket #7608
>
> http://trac.sagemath.org/sage_trac/ticket/7608
>
> R has a third-party package for SNA. The R spkg was recently upgraded
> to version 2.10.1. I don't see what else one could do about R as
> NetworkX is (from my experience) much faster than R's sna package for
> SNA related functionalities (think of networks with tens of thousands
> of nodes). Over the last year, I have come across some feature
> requests on the NetworkX mailing list. With some questioning and
> literature search, it's possible to get an idea of what needs to be
> implemented/improved in NetworkX and/or Sage for SNA.
>
> --
> Regards
> Minh Van Nguyen
Jason Grout
I have a small list of small projects. I started this list for my
students in a numerical analysis class (ideas for their final projects),
but I've added other stuff to it on occasion.
http://orion.math.iastate.edu/grout/sage
Some things on that list are small bug fixes that I should really put up
on trac...
Thanks,
Jason
mhampton
As Mike pointed out, look in sage/interfaces/phc.py for what is
already there. At Jan Verschelde's request I also recently updated
the optional phcpack spkg to a recent version.
The problem with phcpack is that it is mostly written in Ada, and it
seems unlikely in the medium term to become a standard package because
of that. I have made some very minor attempts to write a homotopy
solver in cython but its a very low priority project for me at the
moment.
I would recommend talking to Jan Verschelde and seeing if he has any
suggestions for further work on integrating phcpack and sage. One
flaw right now is that the interface I wrote uses pexpect, instead of
the C interface that Jan and one of his students wrote recently. I'm
not sure that would really make much of an impact in speed though,
since parsing the output is not usually the bottleneck.
Right now the 1-variable case is broken, there is a trac ticket for
that (#4411). I have been lazy about fixing it since there are other
good tools in sage for 1-variable problems. Fixing that would be
pretty easy and give you a better idea of what the current
capabilities are.
There's also this @interact example on the wiki, which shows some of
the graphical possiblities of sage+phcpack:
http://wiki.sagemath.org/interact/algebra#NumericalSolutionsofPolynomialSystemswithPHCpack
-Marshall Hampton
On Feb 25, 12:36 am, Mike Hansen <mhan...@gmail.com> wrote:
> On Wed, Feb 24, 2010 at 9:13 PM, Joshua Herman <zitterbeweg...@gmail.com> wrote:
> > Link tohttp://www.math.uic.edu/~jan/download.html
Dr. David Kirkby
> Yes, I think there should be a Sage GSOC application. I'm OK with
> being listed as a mentor. However, I won't write the actual
> application, since I wrote three applications in a row that were all
> turned down, so clearly I don't know how to write one.
>
> What happened last year, by the way, was that several people wanted to
> write a GSOC application. However, nobody volunteered to mentor any
> Sage-related GSOC projects at all. As a result, we did not even apply
> to have Sage as a mentoring organization.
>
> -- William
>
Perhaps one should consider what fits in best with the Google business model.
How about extending Google's "calculator" to have enhanced functionality like
Wolfram Alpha? i.e replace the calculator with Sage?
Dave
Harald Schilly
> Perhaps one should consider what fits in best with the Google business model.
>
They organize information and make it universally useful. Sage doesn't
help and what you suggest only overlaps with their own search engine
efforts. Their vision is a search engine that gives you instantly the
answer you are looking for and websites only serve as references.
In my eyes, what they try to do with gsoc is a mix of sponsoring open
source projects (because they use them heavily ... i.e. they were
sponsors of the R project long before there was any gsoc at all) and
also do a mix of of marketing+recruiting for potential new employees.
So, what they look for is a project that helps them or a project where
these kind programmers are attracted that they want to have. Hence,
everything with pure mathematics is not on their list [yes, there is
google research, but that's not gsoc] and there are only very few math
related projects that were in gsoc in the past. That's why I already
wrote above that I think that the best chances are with "pure
informatics" related problems where already solutions exist. That must
be communicated in that way, too! I think of things like account
management, user management, interactive websites, compiler
(cython, ...), databases, and probably visualization.
greetings H
Robert Bradshaw
> William Stein wrote:
>
>> Yes, I think there should be a Sage GSOC application. I'm OK with
>> being listed as a mentor. However, I won't write the actual
>> application, since I wrote three applications in a row that were all
>> turned down, so clearly I don't know how to write one.
>> What happened last year, by the way, was that several people wanted
>> to
>> write a GSOC application. However, nobody volunteered to mentor any
>> Sage-related GSOC projects at all. As a result, we did not even
>> apply
>> to have Sage as a mentoring organization.
Of course a huge portion of the community already has their hands full
with full time mentoring (grad students, undergrads, etc).
> Perhaps one should consider what fits in best with the Google
> business model.
>
> How about extending Google's "calculator" to have enhanced
> functionality like Wolfram Alpha? i.e replace the calculator with
> Sage?
If they wanted this they would not be doing this as part of GSoC for
technical, legal, and pragmatic reasons.
GSoC is primarily a philanthropic and marketing program, and though it
of course provides direct benefits to them, they are often of the more
intangible sort.
- Robert