Fwd: LiE questions
David Joyner
This does a number of Lie algebra computations and
representation-theoretic computations which people on this list
have expressed interest in.
---------- Forwarded message ----------
From: Marc van Leeuwen <Marc.van...@math.univ-poitiers.fr>
Date: Aug 24, 2007 5:23 AM
Subject: Re: LiE questions
To: David Joyner <wdjo...@gmail.com>
David Joyner wrote:
> Hi Marc:
>
> I was just thinking about this package again recently for some reason.
>
> I wonder if you could please ask if it is possible for LiE to
> be GPL'd at some point in the future. If not, fine, but I'd
> like to know if you or Prof Cohen think that is likely.
>
> - David
Since the time of that message I have clarified the issue with prof. Cohen.
As is now indicated on the Lie home page
http://www-math.univ-poitiers.fr/~maavl/LiE/
you can distribute LiE under the (GNU) LGPL. It is not clear to me why this
choice was made, since LiE is not distributed as a library, but I think this is
because a part of LiE has thus been incorporated into Magma. In any case, if
you read the fine print of the LGPL, it is allowed to take a copy and
distribute that copy under the GPL (but that copy can then no longer be
converted back to LGPL). I hope this helps you out, best regards,
-- Marc van Leeuwen
Mike Hansen
definitely on my radar. Other packages of interest are
Schur http://schur.sourceforge.net/
Atlas of Lie Groups Software http://www.liegroups.org/software/
--Mike
William Stein
functionality of Lie to SAGE? People have personally requested
this to me about 7 times now.
-- William
--
William Stein
Associate Professor of Mathematics
University of Washington
http://www.williamstein.org
David Joyner
future. I'll actually be meeting de Graaf (who is a GAP developer
and wrote a book on Lie Algebras) in Germany in about
3 weeks and can ask him about it as well.
++++++++++++++++++++++++++++++++++++
Mike Hansen
http://www.sagetrac.org/sage_trac/ticket/489
--Mike
mabshoff
On Aug 24, 6:05 pm, "Mike Hansen" <mhan...@gmail.com> wrote:
> I'll go ahead and do some work on it this weekend. I've created a trac ticket:http://www.sagetrac.org/sage_trac/ticket/489
>
> --Mike
>
<SNIP>
Hello folks,
and it already has been build on neron. Just add -L${SAGE_LOCAL} -
lncurses to the link flags, i.e.
mabshoff@neron LiE$ gcc -o Lie.exe lexer.o parser.o non-ANSI.o
bigint.o binmat.o creatop.o gettype.o getvalue.o init.o learn.o main.o
mem.o node.o onoff.o output.o poly.o sym.o print.o getl.o date.o
static/*.o box/*.o -L /extra/home/mabshoff/SAGE-build/sage-2.8.2.rc1/
local/lib/ -lreadline -lcurses
and it compiles on neron:
mabshoff@neron LiE$ ./Lie.exe
LiE version 2.2.2 created on Aug 24 2007 at 12:52:16
Authors: Arjeh M. Cohen, Marc van Leeuwen, Bert Lisser.
Free source code distribution
type '?help' for help information
type '?' for a list of help entries.
>
I will add the needed bits to the trac ticket. Is there a test suite
for Lie I could/should run?
Cheers,
Michael
David Joyner
In .../LiE/progs there are some test files.
>
> Cheers,
>
> Michael
>
>
> >
>
mabshoff
On Aug 24, 7:18 pm, "David Joyner" <wdjoy...@gmail.com> wrote:
Hello David,
> On 8/24/07, mabshoff <Michael.Absh...@fsmath.mathematik.uni-dortmund.de> wrote:
>
> > I will add the needed bits to the trac ticket. Is there a test suite
> > for Lie I could/should run?
>
> In .../LiE/progs there are some test files.
>
I ran them all and none of them crashed/complained. test1 gave an
awful lot of output, but according to the shell Lie terminated with
exit status 0. Some from the perspective of Solaris there seem to be
no problems which is very good news for the Sage Solaris port, we have
enough unresolved issues as it is there :(. They are getting steadily
fewer, but if there is anybody out there with deep knowledge about the
NTL and gcc 4.2 on Solaris and likes weired linker failures please
speak up.
Is it worth it to give feedback back to the Lie people regarding
Solaris? A quick grep didn't turn up anything useful, but this is the
undocumented source package after all:
mabshoff@neron LiE$ grep -r [Ss]olaris *
Binary file Lie.exe matches
mabshoff@neron LiE$
Cheers,
Michael
David Joyner
>
> Hello David,
>
> > On 8/24/07, mabshoff <Michael.Absh...@fsmath.mathematik.uni-dortmund.de> wrote:
> >
> > > I will add the needed bits to the trac ticket. Is there a test suite
> > > for Lie I could/should run?
> >
> > In .../LiE/progs there are some test files.
> >
>
> I ran them all and none of them crashed/complained. test1 gave an
> awful lot of output, but according to the shell Lie terminated with
> exit status 0. Some from the perspective of Solaris there seem to be
> no problems which is very good news for the Sage Solaris port, we have
> enough unresolved issues as it is there :(. They are getting steadily
> fewer, but if there is anybody out there with deep knowledge about the
> NTL and gcc 4.2 on Solaris and likes weired linker failures please
> speak up.
>
> Is it worth it to give feedback back to the Lie people regarding
> Solaris? A quick grep didn't turn up anything useful, but this is the
Marc van Leeuwen (ma...@mathlabo.univ-poitiers.fr) is the current
maintainer. He seems very helpful and would probably appreciate
feedback.
mabshoff
On Aug 24, 11:15 pm, "David Joyner" <wdjoy...@gmail.com> wrote:
> On 8/24/07, mabshoff <Michael.Absh...@fsmath.mathematik.uni-dortmund.de> wrote:
>
>
>
>
>
> > On Aug 24, 7:18 pm, "David Joyner" <wdjoy...@gmail.com> wrote:
>
> > Hello David,
>
> > > On 8/24/07, mabshoff <Michael.Absh...@fsmath.mathematik.uni-dortmund.de> wrote:
>
> > > > I will add the needed bits to the trac ticket. Is there a test suite
> > > > for Lie I could/should run?
>
> > > In .../LiE/progs there are some test files.
>
> > I ran them all and none of them crashed/complained. test1 gave an
> > awful lot of output, but according to the shell Lie terminated with
> > exit status 0. Some from the perspective of Solaris there seem to be
> > no problems which is very good news for the Sage Solaris port, we have
> > enough unresolved issues as it is there :(. They are getting steadily
> > fewer, but if there is anybody out there with deep knowledge about the
> > NTL and gcc 4.2 on Solaris and likes weired linker failures please
> > speak up.
>
> > Is it worth it to give feedback back to the Lie people regarding
> > Solaris? A quick grep didn't turn up anything useful, but this is the
>
> Marc van Leeuwen (ma...@mathlabo.univ-poitiers.fr) is the current
> maintainer. He seems very helpful and would probably appreciate
> feedback.
>
Great, I am CCing him so he can catch up :-). The specific thread in
the google group sage-devel can be accessed via
http://groups.google.com/group/sage-devel/t/bfa83514467cdfe6
Marc, the Sage people (specifically Mike Hansen) have started writing
a wrapper to call Lie from Sage. We (i.e. me) have also done some
build tests on some common and not so common platforms and found some
issues, mainly the need to link against curses or ncurses. You have
some info about building Lie on OSX at your website and I will
hopefully have some time tomorrow to test it out there.
The is also a small issue in or around Objectwrite that valgrind
exposed. Being unfamiliar with the code somebody more familiar with
the code base might fix it:
==22512== Syscall param write(buf) points to uninitialised byte(s)
==22512== at 0x31058BFA60: __write_nocancel (in /lib64/libc-2.5.so)
==22512== by 0x31058686D2: _IO_file_write@@GLIBC_2.2.5 (in /lib64/
libc-2.5.so)
==22512== by 0x31058685E5: _IO_do_write@@GLIBC_2.2.5 (in /lib64/
libc-2.5.so)
==22512== by 0x310586971F: _IO_file_close_it@@GLIBC_2.2.5 (in /
lib64/libc-2.5.so)
==22512== by 0x310585E249: fclose@@GLIBC_2.2.5 (in /lib64/
libc-2.5.so)
==22512== by 0x40565C: Objectwrite (in /tmp/Work2/LiE/Lie.exe)
==22512== by 0x41241E: vid_writestring_tex_grp_tex (in /tmp/Work2/
LiE/Lie.exe)
==22512== by 0x406325: eval_value (in /tmp/Work2/LiE/Lie.exe)
==22512== by 0x41346F: arg_sequence (in /tmp/Work2/LiE/Lie.exe)
==22512== by 0x406169: eval_value (in /tmp/Work2/LiE/Lie.exe)
==22512== by 0x40291D: yyparse (in /tmp/Work2/LiE/Lie.exe)
==22512== by 0x40764D: main (in /tmp/Work2/LiE/Lie.exe)
==22512== Address 0x4C0B00F is not stack'd, malloc'd or (recently)
free'd
I stuck all the info in the track ticket #489 at http://sagetrac.org/sage_trac/ticket/489
- If you would like a trac account please contact William Stein.I
couldn't find an issue tracker for Lie, if you do have one please let
us know and I will file any new potential issues there.
Let me know if you need any additional info or feedback.
Cheers,
Michael
Mike Hansen
I've created and attached an expect interface for LiE. It has support
for tab-completion, help, etc. I've also attached an initial spkg.
Here's the trac ticket: http://sagetrac.org/sage_trac/ticket/489
Let me know if you have any comments / problems /...
--Mike
On 8/24/07, mabshoff <Michael...@fsmath.mathematik.uni-dortmund.de> wrote:
>
>
>
David Joyner
However, I can't get it to work, nor can I find a
lie.py file in sage/interfaces. Obviously something
is wrong. Can you give a command which works, like
lie.eval("diagram(E8)")
or something pleaaaase?
++++++++++++++++++++++++++++++++++++++=
Mike Hansen
I didn't include the interface with the spkg. It's in the HG bundle.
I've attached the lie.py file. It just needs to put in
sage/interfaces/ . Then, the following line should be added to
sage/interfaces/all.py :
from lie import lie, lie_console, LiE
There is a bit of documentation in lie.py.
Let me know how it goes,
Mike
David Joyner
Looks like you've done a fantastic job.
Thanks very much for doing this!
Mike Hansen
--Mike