On Mar 30, 6:46 pm, William Stein wrote:
> On Mon, Mar 30, 2009 at 8:43 AM, dmitrey wrote:
>
> > Hi there,
> > I've been informed of the discussion,
> > so my 2 cents:
> > OpenOpt can work without CVXOPT installed (you can easily ensure it
> > via running /examples/nlp_1.py, that uses ralg). T
On Mon, Mar 30, 2009 at 8:43 AM, dmitrey wrote:
>
> Hi there,
> I've been informed of the discussion,
> so my 2 cents:
> OpenOpt can work without CVXOPT installed (you can easily ensure it
> via running /examples/nlp_1.py, that uses ralg). The mentioned files
> (http://trac.openopt.org/openopt/br
Hi there,
I've been informed of the discussion,
so my 2 cents:
OpenOpt can work without CVXOPT installed (you can easily ensure it
via running /examples/nlp_1.py, that uses ralg). The mentioned files
(http://trac.openopt.org/openopt/browser/OOPy/openopt/solvers/CVXOPT/
*.py) are imported IF and ON
I just tried again the performances of GLPK against COIN-OR with a new
option I had not noticed before, and the difference is now only of
2 : 1
It is a bit better than the 30 : 1 found with the previous example and
perhaps it will let me code the algorithms I needed without waiting
for COIN-OR to
On Sun, Mar 29, 2009 at 11:56 AM, Harald Schilly
wrote:
>
> On Mar 29, 8:29 pm, William Stein wrote:
>> Since cvxopt is GPLV3+ according to
>>
>> http://abel.ee.ucla.edu/cvxopt/copyright.html
>>
>> OpenOpt can't legally be BSD licensed.
>
> Why not? From what I know, cvxopt has to be installed
On Mar 29, 8:29 pm, William Stein wrote:
> Since cvxopt is GPLV3+ according to
>
> http://abel.ee.ucla.edu/cvxopt/copyright.html
>
> OpenOpt can't legally be BSD licensed.
Why not? From what I know, cvxopt has to be installed separately, not
a part of openopt.
h
--~--~-~--~~---
On Sun, Mar 29, 2009 at 11:18 AM, mabshoff wrote:
>
>
>
> On Mar 29, 10:46 am, William Stein wrote:
>
>
>
>> So CBC can't ever be included in Sage unless CBC changed their license.
>>
>> The only way I see forward to have faster linear programming included
>> *standard* in Sage:
>>
>> 1. Speed
On Sun, Mar 29, 2009 at 11:00 AM, Harald Schilly
wrote:
>
> On Mar 29, 7:46 pm, William Stein wrote:
>> The only way I see forward to have faster linear programming included
>> *standard* in Sage:
>>
>
> Would it be possible to have "some code" that would connect Sage with
> those solvers, but n
On Mar 29, 10:46 am, William Stein wrote:
> So CBC can't ever be included in Sage unless CBC changed their license.
>
> The only way I see forward to have faster linear programming included
> *standard* in Sage:
>
> 1. Speed up GLPK.
> 2. Get the license on CBC changed.
> 3. Find another lin
On Mar 29, 7:46 pm, William Stein wrote:
> The only way I see forward to have faster linear programming included
> *standard* in Sage:
>
Would it be possible to have "some code" that would connect Sage with
those solvers, but not include them? i.e. calling a library and
complaining, if it doesn'
On Sun, Mar 29, 2009 at 7:41 AM, Nathann Cohen wrote:
>
> Hello everybody !!!
>
> I wanted to expand SAGE graph theoretical's functions when I noticed
> that the COIN-OR LP solvers were not even available
>
> I talked about it with a few of you on SAGE-DEVEL Irc channel, and
> ended up writi
11 matches
Mail list logo
