call for success stories

23 views
Skip to first unread message

Harald Schilly

unread,
Jul 15, 2008, 6:15:36 AM7/15/08
to sage-...@googlegroups.com, sage-s...@googlegroups.com, sage...@googlegroups.com
Hi, I'm working on the Sage website.

I am searching for interesting content and this time I want to ask
everyone who has used Sage for his or her research or in education in
class to write a short success story. It should talk about how it was
used and the general and personal experience. Just some sentences to
give new users a first impression of Sage.

thanks and greetings, Harald

mhampton

unread,
Jul 15, 2008, 7:11:27 PM7/15/08
to sage-devel
I've been planning on writing another article on OpenWetWare on my use
of Sage/Cython in modeling cardiovascular systems. I won't have time
to do that for another week or two but I thought I would let you
know. I have another success story with sage in another field but its
harder to do a simple exposition of it (finiteness of polynomial
varieties using techniques from tropical geometry - gfan in sage).

-M. Hampton

Dan Drake

unread,
Jul 15, 2008, 10:25:10 PM7/15/08
to sage-...@googlegroups.com

If you want a sound-bite sized success story:

I needed to do some lengthy calculations using the symbolic math
capabilities of Sage. Using Python's built-in data structures, I could
easily store partial results, which let me stop and restart my
computations. Getting up to speed with Sage was easy since I already knew
Python.

Dan

--
--- Dan Drake <dr...@mathsci.kaist.ac.kr>
----- KAIST Department of Mathematical Sciences
------- http://math.kaist.ac.kr/~drake

signature.asc

tkeller

unread,
Jul 16, 2008, 12:27:47 AM7/16/08
to sage-devel
I'm working on modeling the efficacy of a hypothetical infectious
vaccine (like the live polio vaccine). I have lots of small coding
chunks that makes the sage notebook great for working through each
section of the problem. Sage including scipy and networkx has
eliminated a lot of the tedium I would have had to churn through
otherwise, while native plotting allows easy visualization. An
earlier implementation in Mathematica 6.0 was orders of magnitude
slower for a simpler simulation. While neither implementation was
optimized, my first pass in python/sage was much more effective, and
easier to write and understand.
Thomas
>  signature.asc
> 1KDownload

Simon King

unread,
Jul 16, 2008, 3:37:14 AM7/16/08
to sage-devel
Dear Harald,

On Jul 15, 12:15 pm, "Harald Schilly" <harald.schi...@gmail.com>
wrote:
> I am searching for interesting content and this time I want to ask
> everyone who has used Sage for his or her research or in education in
> class to write a short success story. It should talk about how it was
> used and the general and personal experience. Just some sentences to
> give new users a first impression of Sage.

I hope the following description isn't too long...

Project:
Computation of cohomology rings of finite p-groups with coefficients
in GF(p)

How Sage and its components were used:
- I use Gap to compute certain elementary abelian subgroups.
- Due to Cython, i can use C-programs of David Green for computing
minimal resolutions.
- Cython is a very nice python-like programming language, in which i
created classes for cochains, chain maps, cohomology rings etc.
- In order to find a minimal generating set and a minimal set of
relations for a cohomology ring, i need Gröbner basis computations in
graded commutative rings (or simply "commutative", in characteristic
2); Singular does that job.

Success story:
- Computation time for the cohomology rings of all 267 groups of order
64: About 2h40min in total (on a single CPU).
- Work in progress: Computation of the cohomology rings of groups of
order 128. So far successful for 2221 groups (i.e., 107 are still
missing). Computation time so far: 15 days, parallely on 4 CPUs.

General and personal experience:
- After learning a few days, wrapping the C-programs was surprisingly
easy.
- Writing Cython-code is fun!
- Communication with Singular via an ascii interface was painful until
i found some tricks to reduce the traffic.
- The non-commutative part of Singular has serious bugs. Hence, unless
i find a work around, i can use my programs only in characteristic 2.
- When problems occured, the Sage community was very responsive (Thank
you!)

Cheers
Simon
Reply all
Reply to author
Forward
0 new messages