1 |
On 16:56 Mon 21 Mar , Brian Dolbec wrote: |
2 |
> There is an app called shogun-toolbox http://www.shogun-toolbox.org/ |
3 |
> |
4 |
> that is designed specifically for machine learning. It has many |
5 |
> interface capabilities including python (which they state has the most |
6 |
> complete documentation). There is also many examples for it's use. |
7 |
> |
8 |
> I personally don't know nearly enough about it to know what modeling |
9 |
> methods might be suitable or even if there is one suitable for this |
10 |
> type of model. But I think should be investigated. Implemented this |
11 |
> would have the potential to increase it's success rate over time, |
12 |
> reducing developer load even more. |
13 |
|
14 |
This sounds like something that might be beyond a GSoC project to |
15 |
develop a standalone code base. It would probably work better next year |
16 |
as an enhancement, if this year's project to build an initial working |
17 |
application worked out well. |
18 |
|
19 |
What I would do instead is a "smart" detection using things like |
20 |
flex/bison/pybison. You could first detect the buildsystem type |
21 |
(autotools/cmake/distutils/etc) based on file existence. Then you would |
22 |
parse the build files using the syntax described for your lexical |
23 |
analyzer/parser to build a basic understanding of what build-time |
24 |
options are available and offer them as USE flags. You could also do |
25 |
similar things to detect dependencies by understanding their syntax in |
26 |
the source code. |
27 |
|
28 |
-- |
29 |
Thanks, |
30 |
Donnie |
31 |
|
32 |
Donnie Berkholz |
33 |
Admin, Summer of Code |
34 |
Gentoo Linux |
35 |
Blog: http://dberkholz.com |