On 3/12/2014 11:39 AM, Raffaello Bertini wrote:
Well,
I don't have any particular idea, but I'm interested to study the topics
previously mentioned.
Can I made more then one proposal for the contest?
Yes. we occasionally get students that submit multiple proposals.
Based on past experience, it is best to pick a subject that is of interest and do some reading and research on that. An area that interests you is best because it gives you the motivation to follow through when things get tough and you will get more out of the experience in the end.
Search google for articles on your subject and do a little research. Discuss the ideas you have and make a proposal. We are happy to review student proposals and make suggestions. Many candidates checkout a fork from github and make sure they can compile it and install it locally and read up on the wiki developer howto documents and ask questions.
Hope this helps,
-Steve
On Tue, Mar 11, 2014 at 2:32 PM, Stephen Woodbridge
<woodbri@swoodbridge.com <mailto:woodbri@swoodbridge.com>> wrote:
On 3/11/2014 6:38 AM, Raffaello Bertini wrote:
Hello from Italy, I'm Raffaello Bertini and I'm interested in
GSoc2014. I've got a bachelor degree in computer science and I'm
attending for the master degree, but actually I'm working for a
start-up designing & developing their main system. I'm finishing my
job at the end of April so I've been able to accomplish the
GSoC2014.
I've already worked with GIS and routing algorithm:
* Postrgers/PostGIS * TSP, Dijkstra, ... * pgRouting as a fork.
For GSoc2014 I'm interested in these topics:
* *ASTP*
I saw there are already a pl/pgsql script that solve ATSP by
conversion and Simulated Annealing for solving TSP. I think it's
required to build a specific function inside pgRouting v2 and I have
in mind to develop a 3opt and/or Simulated Annealing for ATSP. Is It
ok?
Yes, the current TSP algorithm only supports symmetric TSP so an
Asymmetric matrix must be converted to a symmetric matrix and then
solved. We think it would be valuable to have an asymmetric solver.
* *Contraction Hierarchies*
I don't understand at all what it is supposed to do. Is it ask to
develop the Contraction Hierarchies algorithm into pgRouting or to
integrate the existing algorithm of OSRM?
Or revising the pgr2OSRM tool? (pgr2OSRM is an utility to connect
pgRouting to OSRM, isn't it?)
We originally thought it would be a good idea to support Contraction
Hierarchies in pgRouting. While this is still and interesting idea, I
needed access to a faster solver and built the osrm-tools to make a
bridge between Project-ORSM and pgRouting
* *Implement generic driving directions add-on to pgRouting*
This one seems very funny to do. I read the wiki "Driving Direction
Instructions".
The challenge here is to create a generic Driving Directions module
that will work with most any edge table and solution. We get regular
requests for help create Driving Directions.
* *Flow Algorithms*
Which one? Maximum or Minimun?
For the Maximum flow
I not familiar with these or what the specific requests have been in
the past.
Thank you for your interest and we look forward to your proposal. We
would also be happy to discuss any idea that you might be interested
in working on.
Best regards,
-Steve
Thanks. Regards Raffaello Bertini
_________________________________________________ pgrouting-dev
mailing
list pgrouting-dev@lists.osgeo.org
<mailto:pgrouting-dev@lists.osgeo.org>
http://lists.osgeo.org/__mailman/listinfo/pgrouting-dev
<http://lists.osgeo.org/mailman/listinfo/pgrouting-dev>
_________________________________________________
pgrouting-dev mailing list
pgrouting-dev@lists.osgeo.org <mailto:pgrouting-dev@lists.osgeo.org>
http://lists.osgeo.org/__mailman/listinfo/pgrouting-dev
<http://lists.osgeo.org/mailman/listinfo/pgrouting-dev>
_______________________________________________
pgrouting-dev mailing list
pgrouting-dev@lists.osgeo.org
http://lists.osgeo.org/mailman/listinfo/pgrouting-dev