You need --math instead of --glp.
Best Regards,
Chris Matrakidis
<http://www.avg.com/email-signature?utm_medium=email_source=link_campaign=sig-email_content=webmail>
Virus-free.www.avg.com
<http://www.avg.com/email-signature?utm_medium=email_source=link_campaign=sig-email_conten
On Mon, 21 Dec 2020 at 01:38, Andrew Makhorin wrote:
> This bug is not fixed yet, because both ssx_chuzc and ssx_chuzr will be
> replaced by new versions.
>
Andrew, what are your plans for the next glpk versions? As you know I have
some stuff that I would like to upstream.
Best Regard
d. For a mip additional preprocessing is possible, a
trivial example is that integer variable bounds can be tightened to the
next integer.
Best Regards,
Chris Matrakidis
f accuracy (attached) and seems to work a lot
better with a subset of miplib problems I used for testing.
Best Regards,
Chris Matrakidis
run_glpk.sh
Description: application/shellscript
Should be fairly easy to do when branching, to me the hard part seems to be
extending the api for this in a clean way.
Best Regards,
Chris Matrakidis
On Mon, 5 Oct 2020 at 16:16, Erwin Kalvelagen <
er...@amsterdamoptimization.com> wrote:
>
> Indicator constraints can be
search for a specific key but only access the previous or next nodes
of a given one.
Best Regards,
Chris Matrakidis
On Sat, 26 Sep 2020 at 19:52, Domingo Alvarez Duarte
wrote:
> Hello Chris !
>
> I've add my fix of avl_tree to your code and compiled it and it segfaults
> due t
gt;
> =
>
> # ubuntu 18.04 glpsol package
>
> INTEGER OPTIMAL SOLUTION FOUND
> Time used: 70.5 secs
> Memory used: 19.8 Mb (20725782 bytes)
> ...
>
> Model has been successfully processed
> 71.49user 0.22system 1:11.71elapsed 99%CPU (0avgtext+0avgdata
> 23500ma
I made some performance improving patches a few years ago. You can find
them in: https://github.com/cmatraki/GLPK
Best Regards,
Chris Matrakidis
On Sat, 26 Sep 2020 at 14:54, Domingo Alvarez Duarte
wrote:
> Hello !
>
> Testing GLPK I left it solving examples/lie_goe.mod for mo
terminating the search instantly, while in the other it takes a few
iterations more (and again terminates the search instantly).
You can't draw any conclusion about solver performance from just one
instance, where you may be lucky (or not).
Best Regards,
Chris Matrakidis
On Tue, 22 Sep 2020 at 20:03
the official releases. There you can also find the
"work" branch with some of my suggested patches.
Best Regards,
Chris Matrakidis
for convenience since I
update it with every new release.
For contributions I send patch files to the list, so that Andrew can pick
them up.
Best Regards,
Chris Matrakidis
___
Help-glpk mailing list
Help-glpk@gnu.org
https://lists.gnu.org/mailman/listinfo/help-glpk
f you
want to take a look.
Best Regards,
Chris Matrakidis
On Tue, 2 Apr 2019 at 15:18, Andrew Makhorin wrote:
> Forwarded Message
> From: Marc Garcia
> To: help-glpk@gnu.org
> Subject: Project development
> Date: Tue, 2 Apr 2019 12:02:16 +0100
>
> Hi ther
constraints helps.
Best Regards,
Chris Matrakidis
___
Help-glpk mailing list
Help-glpk@gnu.org
https://lists.gnu.org/mailman/listinfo/help-glpk
Best Regards,
Chris Matrakidis
___
Help-glpk mailing list
Help-glpk@gnu.org
https://lists.gnu.org/mailman/listinfo/help-glpk
e missing something.
Best Regards,
Chris Matrakidis
___
Help-glpk mailing list
Help-glpk@gnu.org
https://lists.gnu.org/mailman/listinfo/help-glpk
Dear Mike,
The config.h you provided looks correct. Until we find a more permanent
solution, you have two options:
a) run configure with --disable-reentrant, or
b) follow Heinrich's solution
This has been discussed in the past[1] and Andrew seems to favour option a.
Best Regards,
Chris
Dear Mike,
Can you share a copy of the generated config.h?
Best Regards,
Chris Matrakidis
On 15 March 2018 at 15:35, <mike.stegl...@berlin.de> wrote:
> Dear GLPK-Team,
>
>
>
> I have got some problems with building glpk 4.65 on Windows 10 with
> msys/mingw.
ing
the root node less during the search, and consequently requiring fewer
factorisations.
I'm attaching csv files with the results of the runs if you want to see
more details.
Best Regards,
Chris Matrakidis
GLPK 4.65 --cuts --pcost,,,
Name, Dual Bound, Primal Bound, Gap%, Nodes, T
complicated
schemes you are describing. Can you give more details on what you are
observing when the solver reaches the time limit?
Best Regards,
Chris Matrakidis
On 5 March 2018 at 19:05, Pietro Scionti <pietro.scio...@archinet.it> wrote:
> Hello,
> I am solving a MIP problem giving
n solve 36 problems from miplib 2010 within a
time limit of two hours per problem.
Do let me know if you have any comments, suggestions or questions.
Best Regards,
Chris Matrakidis
___
Help-glpk mailing list
Help-glpk@gnu.org
https://lists.gnu.org/mailma
esults are typically obtained if
separation is only invoked at each k-th backtracking (we choose k = 4 in
our implementation)." Indeed, generating cuts after every fourth
backtracking step seems to improve overall solver performance but again I
haven't tested this as much as I would li
ons can be done
to the preprocessed problem (e.g. no column/row deletions). However,
glp_npp_load_sol()
and glp_npp_recover_sol() have exactly the same restrictions, so I don't
see it as a show stopper.
Best Regards,
Chris Matrakidis
___
Help-glpk mailing list
He
oecessed problem using probing
until no further changes are possible; and
b) for additional preprocessing after adding symmetry breaking inequalities
to the preprocessed problem.
Best Regards,
Chris Matrakidis
___
Help-glpk mailing list
Help-glpk@gnu.org
ersion and it seems to be a big improvement. The only issue I run
into was with primal simplex, similar to the case that was reported earlier
[1].
Best Regards,
Chris Matrakidis
[1] https://lists.gnu.org/archive/html/help-glpk/2017-08/msg0.html
___
Hel
own that using
Dijkstra shortest-path method to solve the Lagrangian sub-problem is the
best choice. In general, GUROBI performs a medium level in both algorithms
and SCIPY is always the worst one."
Best Regards,
Chris Matrakidis
___
Help
hanks Rob) and on linux where it
doesn't change anything.
Best Regards,
Chris Matrakidis
mingw.patch
Description: Binary data
___
Help-glpk mailing list
Help-glpk@gnu.org
https://lists.gnu.org/mailman/listinfo/help-glpk
and I would like to isolate the change responsible for this.
Best Regards,
Chris Matrakidis
___
Help-glpk mailing list
Help-glpk@gnu.org
https://lists.gnu.org/mailman/listinfo/help-glpk
hanges in this version that should affect the simplex
method execution. Did you change anything else in your environment
(like C compiler or compilation flags)? Or are you comparing against
an older version of GLPK (and if so which one)?
Best Regards,
Chris Matr
ile?
> Also the interesting part is that on the older version of glpk it was solved
> without any problems.
I tried both version 4.61 and the 4.62 pre-release without problems.
Which one was the older version that run without problems?
Best Regards
t Regards,
Chris Matrakidis
___
Help-glpk mailing list
Help-glpk@gnu.org
https://lists.gnu.org/mailman/listinfo/help-glpk
plex solver when this happens. This further
improves solution times for some instances and allows two more to be
solved (self and stat96v1). I'm not suggesting to add this patch as
is, just sending it for easy testing.
Best Regards,
Chris Matrakidis
diff --git a/src/glpapi06.c b/src/glpapi06.c
index 4a5267c
rary that provides a C interface.
However, when I looked at it a few months ago, the semantics of the
interface looked a bit different to those of the current code so this
was not so easy to test quickly.
Best Regards,
Chris Matrakidis
___
Help-glpk
running intfeas1() on problems without integer objective
coefficients, if no bound is specified. All patches are relative to
version 4.61.
Somewhat related: There are C bindings for the C++ MiniSat [3], so it
may be possible to hook a more recent version to GLPK.
Best Regards,
Chris Matrakidis
PS
Hi Andrew,
The attached patch adds the _s variants for windows. I have only
tested it using mingw64, so some some additional testing would be
appropriate.
Best Regards,
Chris Matrakidis
On 30 January 2017 at 13:07, Andrew Makhorin <m...@gnu.org> wrote:
> Please see an updated versio
tional error
checking for null pointers, which is not an issue in this code.
Best Regards,
Chris Matrakidis
[1] https://msdn.microsoft.com/en-us/library/0z9czt0w(v=vs.80).aspx
___
Help-glpk mailing list
Help-glpk@gnu.org
https://lists.gnu.org/mailman/li
the bfd_copy function that uses them. The
copy operations reuse existing allocations in the destination when
possible.
Best Regards,
Chris Matrakidis
pcost2.patch
Description: Binary data
___
Help-glpk mailing list
Help-glpk@gnu.org
https://lists.gnu
L0 * U0 we should stop.
I suppose this means that a new simplex flag may be needed, to stop
when re-factorisation is required.
Best Regards,
Chris Matrakidis
commit 9734f4815fda152a20491a617818badb631adbf6
Author: Chris Matrakidis <cmatr...@gmail.com>
Date: Sat Jan 14 16:24:45 2017 +0200
ch to implement
the procedure you outlined. If you want, I can make a mock-up using
the bfd_copy() operation which can be later simplified to work with
the Schur complement basis factorisation.
Best Regards,
Chris Matrakidis
___
Help-glpk mai
* U0 is not changed, so B = B0 can be easily restored as
> in case of "eta file".
Is this what you were referring to in
http://lists.gnu.org/archive/html/help-glpk/2012-06/msg00023.html ?
Best Regards,
Chris Matrakidis
___
Help-glpk mai
you think it may be helpful, I
will send it.
Best Regards,
Chris Matrakidis
___
Help-glpk mailing list
Help-glpk@gnu.org
https://lists.gnu.org/mailman/listinfo/help-glpk
,
Chris Matrakidis
diff --git a/src/glpios09.c b/src/glpios09.c
index d87578c..b9e6b5a 100644
--- a/src/glpios09.c
+++ b/src/glpios09.c
@@ -403,6 +403,8 @@ struct csa
double *up_sum; /* double up_sum[1+n]; */
/* up_sum[j] is the sum of per unit degradations of the objective
over
.
The second patch changes two asserts into errors (one in primal and
one in dual). I managed to trigger the second one, but I'm changing
the first one just in case.
Best Regards,
Chris Matrakidis
PS. In addition to the patch [2] you mentioned a few days ago, in May
I sent another patch as well [3
t makes sense to have it enabled by default eventually, but I
think the switch is still necessary to ease configuring a single
threaded version.
Best Regards,
Chris Matrakidis
___
Help-glpk mailing list
Help-glpk@gnu.org
https://lists.gnu.org/mailman/listinfo/help-glpk
on the work of Dmitry Nadezhin [2] and the ax_tls macro
form the autoconf archive [3] simplified to match the configure.ac
style, The code remains C89 if the option is not selected.
I have tested this extensively but only on Linux with gcc, so
additional testing and feedback is welcome.
Best Regards,
Chris
ll.
Calling glp_get_stat() after glp_simplex() returns 0 will give
GLP_UNBND for an unbounded problem.
Best Regards,
Chris Matrakidis
___
Help-glpk mailing list
Help-glpk@gnu.org
https://lists.gnu.org/mailman/listinfo/help-glpk
= GLP_RT_FLIP;
for an LP or
iocp.flip = GLP_ON;
for a MIP.
Other things to try are changing the preprocessing and scaling options.
Do let us know whether any of this helped with your problem.
Best Regards,
Chris Matrakidis
___
Help-glpk mailing list
Help
Lukas,
There is a spelling error in the attached file. The correct should be
"budget: 400x1 +..." and "fte: 0.0x1 +...".
Also note that you don't have to specify terms where the coefficient
is zero, like "0.0x1" above .
B
using --bestp on the command line. However the
proximity search one takes an order of magnitude more iterations for
the first solution. I understand that the search is probably just
taking a different path, but maybe it offers a hint.
Best Regards
Chris Matrakidis
is found quickly with both sse and x87 floating point.
Best Regards,
Chris Matrakidis
___
Help-glpk mailing list
Help-glpk@gnu.org
https://lists.gnu.org/mailman/listinfo/help-glpk
t; 295500: obj = 3.121033256e+008 inf = 5.425e+006 (14)
>> 296000: obj = 3.121033256e+008 inf = 5.425e+006 (14)
>> 296500: obj = 3.121033256e+008 inf = 5.425e+006 (14)
Best Regards,
Chris Matrakidis
___
Help-glpk mailing list
Help-glpk@gnu.org
https://lists.gnu.org/mailman/listinfo/help-glpk
not found yet >= 2.0e+000(2; 0)
+ 10099: mip = not found yet >= 2.0e+000(6; 0)
+ 10131: mip = not found yet >= 2.0e+000(15; 0)
+ 10141: mip = not found yet >= 2.0e+000(24; 0)
...
Best Regards,
Chris Matra
ost initialisation using strong branching.
Best Regards,
Chris Matrakidis
env.patch
Description: Binary data
proxy.patch
Description: Binary data
___
Help-glpk mailing list
Help-glpk@gnu.org
https://lists.gnu.org/mailman/listinfo/help-glpk
ons if necessary.
All four patches apply cleanly to the new version.
Best Regards,
Chris Matrakidis
___
Help-glpk mailing list
Help-glpk@gnu.org
https://lists.gnu.org/mailman/listinfo/help-glpk
t Regards,
Chris Matrakidis
___
Help-glpk mailing list
Help-glpk@gnu.org
https://lists.gnu.org/mailman/listinfo/help-glpk
ards,
Chris Matrakidis
time.patch
Description: Binary data
dual_fix.patch
Description: Binary data
___
Help-glpk mailing list
Help-glpk@gnu.org
https://lists.gnu.org/mailman/listinfo/help-glpk
the basis matrix (1)
Constructing initial basis...
which may not be optimal but I think is acceptable.
Best Regards,
Chris Matrakidis
new_basis.patch
Description: Binary data
___
Help-glpk mailing list
Help-glpk@gnu.org
https://lists.gnu.org/mailman
in a test set of 64 miplib instances).
Best Regards,
Chris Matrakidis
mir.patch
Description: Binary data
___
Help-glpk mailing list
Help-glpk@gnu.org
https://lists.gnu.org/mailman/listinfo/help-glpk
Andrew,
The attached patch adds a tan() function to MathProg. I was quite surprised
when I tried to use it and got an error message, so I decided to implement
it.
Best Regards,
Chris Matrakidis
tan.patch
Description: Binary data
___
Help-glpk
lower (or upper) limit. Therefore the solution is certain
to be at most 20% off the optimum.
Best Regards,
Chris Matrakidis
___
Help-glpk mailing list
Help-glpk@gnu.org
https://lists.gnu.org/mailman/listinfo/help-glpk
provide an initial solution I find even more
awkward... You call glp_intopt(), with presolving disabled, with a callback
that calls glp_ios_heur_sol() followed by glp_ios_terminate(). Then you can
call glp_intopt() a second time with presolving enabled and the use_sol
parameter set.
Best Regards,
not be correct for the preprocessed problem until
a new one is found.
Best Regards,
Chris Matrakidis
___
Help-glpk mailing list
Help-glpk@gnu.org
https://lists.gnu.org/mailman/listinfo/help-glpk
le together
with a callback.
My preference is to have a flag in glp_iocp to enable the rounding
heuristic (like fp_heur and ps_heur) that would be enabled by
glp_init_iocp() but the user can then disable. Alternatively (or perhaps
additionally), there could be a
in case
anyone gets an "Assertion failed: kk <= n" error and wants to check if this
is the reason, I would advise against using it for any other reason.
Best Regards,
Chris Matrakidis
--- main.c.org 2015-10-15 10:00:00.0 +0300
+++ main.c 2015-10-19 00:46:37.5 +03
63 matches
Mail list logo