[SCIP] SCIP Optimization Suite 3.1.0
Gregor Hendel
hendel at zib.de
Mon Apr 8 01:12:05 CEST 2019
Dear Leila,
Am 01.04.19 um 14:42 schrieb ghasemzadeh at itu.edu.tr:
>
> Dear Hendel,
>
> Thank you so much for your prompt and detailed response.
>
> I prefer to implement the method in SCIP 6.0.1 but the available code
> can be run just in SCIP 3.1.0. I contacted the author and asked for a
> newer version of their implementation but they didn't have.
>
I didn't expect that. But what did they answer when you asked the
question "I am trying to compile your package, but I get an error. Here
is my setup. Can you help?" Infact, if they provide you with source
code, it is their responsibility to also provide you with a README file
on how to set things up. I assume that on this entire mailing list with
400+ people, nobody can do that because nobody has the source code. If
they don't have a README, ask for it. If they provide a README, but you
struggle following the steps, ask them for clarification.
>
>
> 1) If you have the implementation of the code in SCIP6, it would be so
> helpful if I can have it.
>
(Unfortunately?) no.
>
>
> 2) For the installation error I provided in my previous email, my
> system config is as follows:
> OS: Ubuntu 18.04
> Compiler: gcc 7.3.0
> I used scipoptsuite-3.1.0
>
> Indeed, The scipoptsuite suite is installed and tested completely and
> once I try to make the paper code implementation, it faces the
> problems which are screenshot in my previous email.
>
> 3) BTW, as i understood it is better to use SCIP and CPLEX (as the LP
> solver) to implement that code instead of using scipoptsuit, right?
>
It depends. I suggest to reread my answer below.
Good luck,
Gregor
>
>
>
> waiting for your response,
>
> All the best,
> Leila
>
>
>
>
>
> Alıntı Gregor Hendel <hendel at z <mailto:hendel at z>ib.de>:
>
>> Good morning Leila,
>>
>> welcome to the SCIP mailing list. I compiled the code by Hal and He a
>> while ago. As you can imagine, I do not recall every detail of the
>> procedure, but He was very helpful and responsive in sharing her
>> setup with me. I answer your questions inline.
>>
>> Side note. I would not recommend to rely on SCIP 3.1.0, which is very
>> outdated compared to SCIP 6.0.1. But, of course, you can start by
>> compiling the original setup first, and then discussing an update
>> with your supervisor.
>>
>>> 1) First of all, I want to know that what is the difference between
>>> SCIP and SCIP Optimization Suite? I searched but did not find any
>>> answer for it.
>> SCIP itself is the main tool you should work with, for setup reasons.
>> I remember that the authors used IBM CPLEX as LP solver. The SCIP
>> Optimization Suite comprises SCIP and additional tools such as the LP
>> solver SoPlex and the modeling language ZIMPL, which you are welcome
>> to try out, but may not be necessary for your concrete project. If,
>> however, you do not have a CPLEX installation at the faculty, use the
>> Optimization Suite instead.
>>> 2) To install the version 3.1.0 should I install both SCIP and SCIP
>>> Optimization Suite? or just one of them is enough?
>> See 1)
>>> 3) I could install sipoptsuite, but for installation of scip I face
>>> some problems. Once, I want to compile another package
>>> (above-mentioned paper) I face some errors, which is attached as
>>> screenshots. It seems that it cannot perform SCIP_CALL funcitons.
>> Probably, you should first ask for help from the original authors of
>> the dagger package. Presumably, they compiled it successfully once.
>> The problem stems from linker errors. In order to help you with that,
>> you must include more information about your setup (Operating System,
>> Compiler Version, package build system).
>>
>> Let us know if they responded and helped you out,
>> Gregor
>>>
>>> _______________________________________________
>>> Scip mailing list
>>> Scip at zib.de
>>> https://listserv.zib.de/mailman/listinfo/scip
>>>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.zib.de/pipermail/scip/attachments/20190408/bc4260df/attachment.html>
More information about the Scip
mailing list