<html>
<head>
<meta content="text/html; charset=ISO-8859-1"
http-equiv="Content-Type">
</head>
<body text="#000000" bgcolor="#FFFFFF">
Dear Mahdi,<br>
<br>
if you do not know how to use a method or don't know its parameters,
check the documentation: <a class="moz-txt-link-freetext" href="http://scip.zib.de/doc/html/">http://scip.zib.de/doc/html/</a><br>
You can also search methods by name there.<br>
<br>
What I wanted you to do is check the *old* variable for whether it
is in the LP and has negative reduced costs. Also you could check
its value in the LP.<br>
<br>
But I have two other things:<br>
1) Do you add the variables to the original or the transformed
constraints? -> SCIPconsIsTransformed()<br>
2) You have upper bounds on your variables. This can be a problem
for column generation, because a variable can have negative reduced
cost if it is nonbasic on its upper bound. Please change it to
infinity or set the upper bound to be lazy (if it is enforced by
some constraint). See the FAQ about how to do this.<br>
<br>
Best,<br>
Gerald<br>
<div class="moz-cite-prefix"><br>
</div>
<blockquote
cite="mid:CAH4sNwq7bkMQU38oK23k+SWNEatFDk5MoZ==L-yZp9p4BcxSPw@mail.gmail.com"
type="cite">
<div dir="ltr">
<div class="gmail_default"
style="font-family:georgia,serif;color:rgb(32,18,77)">>>is
the delay flag of your pricer set to TRUE? <br>
</div>
<div class="gmail_default"
style="font-family:georgia,serif;color:rgb(32,18,77)">
Yes,<br>
<br>
</div>
<div class="gmail_default"
style="font-family:georgia,serif;color:rgb(32,18,77)"><br>
>>When you create and add new variables, do you set the
initial flag of the variables to TRUE? <br>
Which parameter corresponds to the initial flag?<br>
<br>
>>You add them with SCIPaddPricedVar()? <br>
Yes,<br>
>>Can you double-check, that the variables you create
really have the same meaning.<br>
Yes,<br>
<br>
>> If so, please check the following for the "old"
variable:<br>
>>- is it in the LP? -> SCIPvarIsInLP()<br>
</div>
<div class="gmail_default"
style="font-family:georgia,serif;color:rgb(32,18,77)">I did
not use SCIPvarIsInLP() ( I do not know how to use it) but in
each iteration I checked, s<br>
</div>
<div class="gmail_default"
style="font-family:georgia,serif;color:rgb(32,18,77)"><br>
>>- what are the reduced cost of the variable ->
SCIPgetVarRedcost()<br>
>>- does this corresond to the reduced cost you are
computing for the new variable?<br>
</div>
<div class="gmail_default"
style="font-family:georgia,serif;color:rgb(32,18,77)">When I
add SCIPgetVarRedCost(), after one iteration it returns error:
<br>
ERROR: Error <1516178640> in function call<br>
</div>
<div class="gmail_default"
style="font-family:georgia,serif;color:rgb(32,18,77)">But in
this iteraion, the reduced cost are exactly the same as what I
calculate.<br>
</div>
<div class="gmail_default"
style="font-family:georgia,serif;color:rgb(32,18,77)">
How can I find out what is wrong here?<br>
Note that without SCIPgetVarRedCost(), I do not get any error.<br>
</div>
<div class="gmail_default"
style="font-family:georgia,serif;color:rgb(32,18,77)">I have
attached the procedure in which I add new variables.<br>
<br>
</div>
<div class="gmail_default"
style="font-family:georgia,serif;color:rgb(32,18,77)">Regards,<br>
Mahdi<br>
</div>
</div>
<div class="gmail_extra"><br>
<br>
<div class="gmail_quote">On 1 April 2014 14:35, Gerald Gamrath <span
dir="ltr"><<a moz-do-not-send="true"
href="mailto:gamrath@zib.de" target="_blank">gamrath@zib.de</a>></span>
wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0
.8ex;border-left:1px #ccc solid;padding-left:1ex">
<div bgcolor="#FFFFFF" text="#000000">
<div>Hi Mahdi,<br>
<br>
is the delay flag of your pricer set to TRUE? When you
create and add new variables, do you set the initial
flag of the variables to TRUE? You add them with
SCIPaddPricedVar()? Can you double-check, that the
variables you create really have the same meaning. If
so, please check the following for the "old" variable:<br>
- is it in the LP? -> SCIPvarIsInLP()<br>
- what are the reduced cost of the variable ->
SCIPgetVarRedcost()<br>
- does this corresond to the reduced cost you are
computing for the new variable?<br>
<br>
Best,<br>
Gerald
<div>
<div class="h5"><br>
<br>
On 01.04.2014 14:51, mahdi noorizadegan wrote:<br>
</div>
</div>
</div>
<div>
<div class="h5">
<blockquote type="cite">
<div dir="ltr">
<div class="gmail_default"
style="font-family:georgia,serif;color:rgb(32,18,77)">Hi
Gerald,<br>
<br>
</div>
<div class="gmail_default"
style="font-family:georgia,serif;color:rgb(32,18,77)">Please
find the attachments.<br>
<br>
</div>
<div class="gmail_default"
style="font-family:georgia,serif;color:rgb(32,18,77)">Thanks,<br>
Mahdi<br>
</div>
</div>
<div class="gmail_extra"><br>
<br>
<div class="gmail_quote">On 1 April 2014 13:30,
Gerald Gamrath <span dir="ltr"><<a
moz-do-not-send="true"
href="mailto:gamrath@zib.de" target="_blank">gamrath@zib.de</a>></span>
wrote:<br>
<blockquote class="gmail_quote" style="margin:0
0 0 .8ex;border-left:1px #ccc
solid;padding-left:1ex">
<div bgcolor="#FFFFFF" text="#000000">
<div>Hi Mahdi,<br>
<br>
Nalan sent these log files already. You
can print statistics by "disp stat" in the
interactive shell or calling
SCIPprintStatistics().<br>
<br>
Best,<br>
Gerald
<div>
<div><br>
<br>
<br>
On 01.04.2014 13:42, mahdi
noorizadegan wrote:<br>
</div>
</div>
</div>
<div>
<div>
<blockquote type="cite">
<div dir="ltr">
<div class="gmail_default"
style="font-family:georgia,serif;color:rgb(32,18,77)">Hi
Gerald,<br>
<br>
</div>
<div class="gmail_default"
style="font-family:georgia,serif;color:rgb(32,18,77)">The
constraints are modifiable. <br>
</div>
<div class="gmail_default"
style="font-family:georgia,serif;color:rgb(32,18,77)">I
am nit sure if you have the whole
log files.<br>
</div>
<div class="gmail_default"
style="font-family:georgia,serif;color:rgb(32,18,77)">
I have attached again the log
files.<br>
</div>
<div class="gmail_default"
style="font-family:georgia,serif;color:rgb(32,18,77)">I
was wondering if you could let me
know what you mean by printing the
statistics after stopping?<br>
</div>
<div class="gmail_default"
style="font-family:georgia,serif;color:rgb(32,18,77)">What
do you want to see in the log
files?<br>
<br>
</div>
<div class="gmail_default"
style="font-family:georgia,serif;color:rgb(32,18,77)">
Regards,<br>
Mahdi<br>
</div>
</div>
<div class="gmail_extra"><br>
<br>
<div class="gmail_quote">On 1 April
2014 10:04, Gerald Gamrath <span
dir="ltr"><<a
moz-do-not-send="true"
href="mailto:gamrath@zib.de"
target="_blank">gamrath@zib.de</a>></span>
wrote:<br>
<blockquote class="gmail_quote"
style="margin:0 0 0
.8ex;border-left:1px #ccc
solid;padding-left:1ex">
<div bgcolor="#FFFFFF"
text="#000000">
<div>Dear Nalan,<br>
<br>
Mahdi's mail didn't reach
the mailing list, because he
also sent it from a mail
address not registered for
the list.<br>
<br>
It seems like your problem
is not caused by the
branching, on two of the
three log file, you already
stop before the first
branching is performed.<br>
<br>
If you could print the
statistics after stopping
and send the log files
again, this would help us a
lot to identify the problem.<br>
<br>
Did you change all
constraints to modifiable to
which you add variables
during solving? Otherwise,
propagation might fix some
of your priced variables to
0 and it might happen that
pricing regenerates them
again and again. Except for
that, my best guess would be
that you somehow compute the
reduced cost wrongly.<br>
<br>
Best,<br>
Gerald
<div>
<div><br>
<br>
<br>
On 01.04.2014 09:20,
Gulpinar, Nalan wrote:<br>
</div>
</div>
</div>
<div>
<div>
<blockquote type="cite">
<div
style="font-size:12pt;font-family:Calibri,Arial,Helvetica,sans-serif">
<p><span
style="font-size:12pt">Dear
Ambros, </span><br>
</p>
<div
style="color:#282828">
<div>
<div
style="color:#000000;font-size:12pt;font-family:calibri,arial,helvetica,sans-serif">
<p><br>
</p>
<p>Thank you for
your prompt
reply. As you
suggested, I
am resending
my previous
email. <br>
</p>
<p><br>
</p>
<p>I am writing
this
email about
the problem we
have faced
with SCIP. <br>
</p>
<p><span
style="font-size:12pt">As Mahdi
reported
yesterday
(please </span><span
style="font-size:12pt">s</span><span style="font-size:12pt">ee his
email </span><span
style="font-size:12pt">b</span><span style="font-size:12pt">elow), we
have checked
the </span><br>
</p>
<p>code several <span
style="font-size:12pt">times, and even t</span><span
style="font-size:12pt">rie</span><span
style="font-size:12pt">d</span><span style="font-size:12pt"> to run its
alternative
formulation
but the
problem </span></p>
<p><span
style="font-size:12pt"></span><span
style="font-size:12pt">remains the </span><span style="font-size:12pt">same.
I have also
attached the
output of
three
instances of
the model. </span></p>
<p><span
style="font-size:12pt">I
hope </span><span
style="font-size:12pt">this helps you to identify the problem. </span></p>
<p><span
style="font-size:12pt"><br>
We use column
generation
procedure
where
continuous
variables are
added into
the </span></p>
<p><span
style="font-size:12pt">master
problem and
branch and
bound
procedure is
used to solve
the master
problem. </span></p>
<p><span
style="font-size:12pt">But
after some
iterations it
cannot be
improved. For
example, as
you will see
from </span></p>
<p><span
style="font-size:12pt">log
file.rtf, we
have the
following output. </span></p>
<p><span
style="font-size:12pt"><br>
</span></p>
<p
style="font-family:calibri,arial,helvetica,sans-serif;font-size:16px"><span
style="font-size:12pt">I am looking forward to receiving your reply and
suggestions. </span></p>
<p
style="font-family:calibri,arial,helvetica,sans-serif;font-size:16px"><span
style="font-size:12pt"><br>
</span></p>
<p
style="font-family:calibri,arial,helvetica,sans-serif;font-size:16px"><span
style="font-size:12pt">Thank you for your help at advance. </span></p>
<p
style="font-family:calibri,arial,helvetica,sans-serif;font-size:16px"><span
style="font-size:12pt"><br>
</span></p>
<p
style="font-family:calibri,arial,helvetica,sans-serif;font-size:16px"><span
style="font-size:12pt">Best regards</span></p>
<p
style="font-family:calibri,arial,helvetica,sans-serif;font-size:16px"><span
style="font-size:12pt"><br>
</span></p>
<p
style="font-family:calibri,arial,helvetica,sans-serif;font-size:16px"><span
style="font-size:12pt">Nalan </span></p>
<div
style="font-family:calibri,arial,helvetica,sans-serif;font-size:16px">
<p><br>
</p>
<div
style="font-family:tahoma;font-size:13px">_____________________________________________<br>
Nalan
Gulpinar,
Ph.D. <br>
Associate
Professor of
Operational
Research<br>
Warwick
Business
School<br>
The University
of Warwick,
Coventry, CV4
7AL, UK.<br>
Tel: <a
moz-do-not-send="true"
href="tel:%2B44%20%280%29%2024%207652%204491" value="+442476524491"
target="_blank">+44
(0) 24 7652
4491</a><br>
Fax: <a
moz-do-not-send="true"
href="tel:%2B44%20%280%29%2024%207652%204539" value="+442476524539"
target="_blank">+44
(0) 24 7652
4539</a><br>
Email: <a
moz-do-not-send="true"
href="mailto:Nalan.Gulpinar@wbs.ac.uk" target="_blank">Nalan.Gulpinar@wbs.ac.uk</a><br>
_____________________________________________<br>
</div>
</div>
<p><span
style="font-size:12pt"><br>
</span></p>
<p><span
style="color:#20124d;font-family:georgia,serif">leastJ/leastT=
0,0,
minRedCost:
-43.3</span><span
style="color:#20124d;font-family:georgia,serif">117</span><span
style="font-size:12pt"><br>
</span></p>
<div
style="color:#20124d;font-size:16px;font-family:georgia,serif">
<p>total cost=
1173.06,
10.84, 882.22,
280, z_134</p>
<p>134, add
new variables</p>
<p> 0.1s|
1 | 0 |
203 | - |
725k| 0 |
33 | 206 | 115
| 205 | 115
| 0 | 0
| 0 |
-- |
7.654555e+03
| Inf </p>
<p> time |
node | left
|LP iter|LP
it/n| mem
|mdpt |frac
|vars |cons
|cols |rows
|cuts
|confs|strbr|
dualbound |
primalbound
| gap </p>
<p> 0.1s|
1 | 0 |
205 | - |
725k| 0 |
36 | 206 | 115
| 206 | 115
| 0 | 0
| 0 |
-- |
7.654555e+03
| Inf </p>
<p> </p>
<p> leastJ/leastT=
2,0,
minRedCost:
-90.6829</p>
<p>total cost=
532.45, 1.05,
363.4, 168,
z_135</p>
<p>135, add
new variables</p>
<p> 0.1s|
1 | 0 |
205 | - |
727k| 0 |
36 | 207 | 115
| 206 | 115
| 0 | 0
| 0 |
-- |
7.654555e+03
| Inf </p>
<p> 0.1s|
1 | 0 |
205 | - |
727k| 0 |
23 | 207 | 115
| 207 | 115
| 0 | 0
| 0 |
-- |
7.654555e+03
| Inf </p>
<p> </p>
<p> leastJ/leastT=
2,0,
minRedCost:
-90.6829</p>
<p>total cost=
532.45, 1.05,
363.4, 168,
z_136</p>
<p>136, add
new variables</p>
<p> 0.1s|
1 | 0 |
205 | - |
729k| 0 |
23 | 208 | 115
| 207 | 115
| 0 | 0
| 0 |
-- |
7.654555e+03
| Inf </p>
<p> 0.1s|
1 | 0 |
209 | - |
729k| 0 |
23 | 208 | 115
| 208 | 115
| 0 | 0
| 0 |
-- |
7.654555e+03
| Inf </p>
<p> </p>
<p> leastJ/leastT=
1,0,
minRedCost:
-88.4043</p>
<p>total cost=
350.84, 31.14,
151.7, 168,
z_137</p>
<p>137, add
new variables</p>
<p> 0.1s|
1 | 0 |
209 | - |
731k| 0 |
23 | 209 | 115
| 208 | 115
| 0 | 0
| 0 |
-- |
7.654555e+03
| Inf </p>
<p> </p>
<p> leastJ/leastT=
1,0,
minRedCost:
-88.4043</p>
<p>total cost=
350.84, 31.14,
151.7, 168,
z_138</p>
<p>138, add
new variables</p>
<p> 0.1s|
1 | 0 |
209 | - |
733k| 0 |
23 | 210 | 115
| 208 | 115
| 0 | 0
| 0 |
-- |
7.654555e+03
| Inf </p>
<p> </p>
<p> leastJ/leastT=
1,0,
minRedCost:
-88.4043</p>
<p>total cost=
350.84, 31.14,
151.7, 168,
z_139</p>
<p>139, add
new variables<br>
</p>
<br>
This goes
on......<br>
</div>
<div
style="color:#282828">
<div dir="ltr">
<div> <br>
</div>
</div>
<div>
<div dir="ltr">
<div
class="gmail_quote">
<div>
<div>
<div dir="ltr"><br>
<div
class="gmail_quote">----------
Forwarded
message
----------<br>
From: <b
class="gmail_sendername">mahdi
noorizadegan</b>
<span
dir="ltr"><<a
moz-do-not-send="true" href="mailto:m.noorizadegan@gmail.com"
target="_blank">m.noorizadegan@gmail.com</a>></span><br>
Date: 31 March
2014 18:37<br>
Subject:
Column
generation<br>
To: <a
moz-do-not-send="true"
href="mailto:Scip@zib.de" target="_blank">Scip@zib.de</a><br>
<div dir="ltr">
<div
style="font-family:georgia,serif;color:#20124d"><br>
</div>
<div
style="font-family:georgia,serif;color:#20124d">Hi,<br>
<br>
</div>
<div
style="font-family:georgia,serif;color:#20124d">I
have been
trying to
solve a
problem by the
column
generation
method.<br>
</div>
<div
style="font-family:georgia,serif;color:#20124d">The
variables
which are
being added to
the master
problem were
binary<br>
</div>
<div
style="font-family:georgia,serif;color:#20124d">so
I had the
problem with
the branch and
bound
procedure.<br>
</div>
<div
style="font-family:georgia,serif;color:#20124d">I
changed the
formulation
and added new
dummy binary
variables and
changed<br>
</div>
<div
style="font-family:georgia,serif;color:#20124d">the
type of the
other
variables to
continuous.<br>
</div>
<div
style="font-family:georgia,serif;color:#20124d">So
there should
not be any
problem with
the branch and
bound
producer.<br>
</div>
<div
style="font-family:georgia,serif;color:#20124d">However,
I still have
problems. It
keeps adding
the same
column after
few
iterations!<br>
</div>
<div
style="font-family:georgia,serif;color:#20124d">For
example for a
simple
instance, the
objective
function value
is 350.84,<br>
</div>
<div
style="font-family:georgia,serif;color:#20124d">the
dual value is
439.24
therefore the
reduced cost
is
-439.24+350.84=-88.40.<br>
</div>
<div
style="font-family:georgia,serif;color:#20124d">When
this new
variable is
added to the
problem, in
the next
iteration this
variable is
again <br>
</div>
<div
style="font-family:georgia,serif;color:#20124d">identified
and added!<br>
<br>
</div>
<div
style="font-family:georgia,serif;color:#20124d">I
have checked
the pricing
problem. It
seems correct
and I could
not find any
mistake.<br>
</div>
<div
style="font-family:georgia,serif;color:#20124d">I
was wondering
if you could
help me and
let me know
what can be
wrong!?<br>
<br>
</div>
<div
style="font-family:georgia,serif;color:#20124d">Thanks
and best,<br>
Mahdi<br>
</div>
</div>
</div>
<br>
</div>
</div>
</div>
</div>
<br>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
</blockquote>
<br>
</div>
</div>
</div>
<br>
_______________________________________________<br>
Scip mailing list<br>
<a moz-do-not-send="true"
href="mailto:Scip@zib.de"
target="_blank">Scip@zib.de</a><br>
<a moz-do-not-send="true"
href="http://listserv.zib.de/mailman/listinfo/scip"
target="_blank">http://listserv.zib.de/mailman/listinfo/scip</a><br>
<br>
</blockquote>
</div>
<br>
</div>
</blockquote>
<br>
</div>
</div>
</div>
</blockquote>
</div>
<br>
</div>
</blockquote>
<br>
</div>
</div>
</div>
</blockquote>
</div>
<br>
</div>
</blockquote>
<br>
</body>
</html>