[GSoC] A proposal for GSoC regarding computational algebra
Hello, cafe I am a currently an undergraduate student and will enter the graduate school on April. As a student, I have a project to propose for the GSoC: purely functional computer algebra library. I've been working on computational-algebra library [1], but this library needs more improvements. Especially, it needs much more performance improvements for Groebner bases computation. I want to implement efficient algorithms called F4 and F5 [2, 3] as the GSoC project. These algorithms require efficient linear algebra algorithms, so this project also involves the development of efficient symbolic linear algebra library. This project is not on any ideas list as far as I know, and I want to directly propose to the GSoC and I'm looking for a mentor. Is there anyone interested in my proposal? [1]: http://github.com/konn/computational-algebra [2]: http://www-polsys.lip6.fr/~jcf/Papers/F99a.pdf [3]: http://www.risc.jku.at/Groebner-Bases-Bibliography/gbbib_files/publication_5... -- Hiromi ISHII konn.jinro@gmail.com
I suppose given your library's use of my algebra package and other
components under the hood, etc. I'd probably be the most likely mentor.
I'd be willing to work with you, and likely to fold it in/replace much of
the existing algebra machinery, which I confess is woefully
under-maintained.
Usually, proposals to write a new library have a hard time getting accepted
to GSOC, but you do already have a decent sized body of work there.
-Edward
On Mon, Mar 17, 2014 at 10:11 AM, Hiromi ISHII
Hello, cafe
I am a currently an undergraduate student and will enter the graduate school on April. As a student, I have a project to propose for the GSoC: purely functional computer algebra library.
I've been working on computational-algebra library [1], but this library needs more improvements. Especially, it needs much more performance improvements for Groebner bases computation. I want to implement efficient algorithms called F4 and F5 [2, 3] as the GSoC project. These algorithms require efficient linear algebra algorithms, so this project also involves the development of efficient symbolic linear algebra library.
This project is not on any ideas list as far as I know, and I want to directly propose to the GSoC and I'm looking for a mentor. Is there anyone interested in my proposal?
[1]: http://github.com/konn/computational-algebra [2]: http://www-polsys.lip6.fr/~jcf/Papers/F99a.pdf [3]: http://www.risc.jku.at/Groebner-Bases-Bibliography/gbbib_files/publication_5...
-- Hiromi ISHII konn.jinro@gmail.com
_______________________________________________ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Hiromi ISHII
Is there anyone interested in my proposal?
I might be. Can you give some use cases? I should point out I do a fair amount of numerical work and this typically involves linear algebra but from a numerical point of view. I am also quite interested in computational algebraic topology but I suspect you are not proposing to work on that?
Hello cafe, I'm so happy to know that there are people interetested in my proposal! I really thank carter (in freenode chat), Edward and Dominic in advance. Edward:
I suppose given your library's use of my algebra package and other components under the hood, etc. I'd probably be the most likely mentor.
I'd be willing to work with you, and likely to fold it in/replace much of the existing algebra machinery, which I confess is woefully under-maintained. Thanks! `algebra` packge provides fine-grained abstraction for algebraic structures (though it does not provide the class for noetherian rings), so I adopted for my purpose. I suppose that the `algebra` is more general purpose library than my `computational-algebra`, which is currently concentrated on computation in polynomial rings or quotient ring.
Usually, proposals to write a new library have a hard time getting accepted to GSOC, but you do already have a decent sized body of work there.
Sounds great. Yes, in fact, my proposal is not building new library but to improve (my personal) existing library. Dominic:
I might be. Can you give some use cases? The current applications of my interest are elimination theory and solving multivariate nonlinear equation systems. Here are some example: https://github.com/konn/computational-algebra/blob/master/examples/solve.hs Another example is purely mathematical things: for example, we can calculate ideal operations in polynomial rings and basic operation in quotient ring. There are application also in the area of statistics, robotics and cryptology, but I don't know much about them.
I should point out I do a fair amount of numerical work and this typically involves linear algebra but from a numerical point of view. That sounds interesting. Linear computations required by F4 and F5 algorithm is purely symbolic ones, but there might be common technique.
I am also quite interested in computational algebraic topology but I suspect you are not proposing to work on that? Sadly not. I don't know much about *algebraic topology*, but my project can be applied to *algebraic geometry* because some computations in commutative algebra can be done with Groebner basis.
-- Hiromi ISHII konn.jinro@gmail.com
participants (3)
-
Dominic Steinitz -
Edward Kmett -
Hiromi ISHII