timelimit option
Answered
I am attempting to set a time limit in AMPL using the following code:
option gurobi_options 'timelimit=7200 outlev=1 mipgap=0 mipfocus=3';
Unfortunately, it does not seem to be functioning as expected. I have over 100 instances, and my goal is to establish a solve time limit of 2 hours for each instance. Could you please let me know why this might not be working?
-
The code snippet you've provided seems correct for setting a time limit when using Gurobi with AMPL. The option
timelimit=7200should indeed set a limit of 7200 seconds (2 hours) for each optimization instance. If it's not functioning as expected, there are a few potential issues to consider:-
Instance Reset: Ensure that the time limit is being correctly applied to each individual instance. If you're running multiple instances in a loop or batch, the time limit needs to be reset or reaffirmed for each instance.
-
AMPL Environment: Double-check the environment in which you are running AMPL. Make sure that the version of AMPL and Gurobi you're using are compatible and correctly configured.
-
Gurobi License and Configuration: Verify that your Gurobi license is valid and that there are no other configuration issues that might be overriding or affecting the
timelimitsetting. -
Problem Complexity: Consider the complexity of the instances. If the problems are very simple, Gurobi might solve them so quickly that the time limit does not come into play. Conversely, if the problems are extremely complex, Gurobi might reach a stopping criterion other than the time limit (like reaching the specified mipgap or due to other limits).
-
Gurobi Log: Check the Gurobi log for each instance to see if there are any messages related to the time limit. The log might provide clues about whether the time limit is being recognized and what's happening when it's reached.
0 -
-
That is a correct way to set the Gurobi time limit, and it's verified to be working with Gurobi 11.0 for AMPL. To be sure you are using it correctly, look for timelimit=7200 or lim:time=7200 in AMPL's Gurobi log output. If you don't see one of these, you can display the actual value of gurobi_options by running again with the AMPL command "option gurobi_options;" right before "solve;".
If the setting of the time limit checks out, then can you explain in more detail how the Gurobi time limit is functioning, and why that's not as expected?
0 -
Thank you so much for the quick response.
I used the following code:option gurobi_options 'timelimit=100 outlev=1 mipgap=0 mipfocus=3';
option gurobi_options 'logfile=C:\AMPL\2023_01_29\Replication01\logfile.temp';
option log_file 'C:\AMPL\2023_01_29\Replication01\logfile.temp';
solve;I set the time limit to 100 seconds for testing purposes, but the solver continued running beyond that limit, forcing me to interrupt it. Therefore, it is still not functioning as expected. Please let me know if there are any corrections needed in the code.
The following is the information from the log file:ampl: model C:\AMPL\2023_01_29\Replication01\2024_03_03_ModelFile_No_Loop.mod;
Gurobi 11.0.0 (win64) logging started Wed Mar 6 17:59:19 2024
Set parameter LogFile to value "C:\AMPL\2023_01_29\Replication01\logfile.temp"
Set parameter InfUnbdInfo to value 1
Gurobi Optimizer version 11.0.0 build v11.0.0rc2 (win64 - Windows 10.0 (19045.2))
CPU model: Intel(R) Core(TM) i7-6700 CPU @ 3.40GHz, instruction set [SSE2|AVX|AVX2]
Thread count: 4 physical cores, 8 logical processors, using up to 8 threads
Optimize a model with 5620 rows, 3051 columns and 136173 nonzeros
Model fingerprint: 0xbac440c9
Variable types: 201 continuous, 2850 integer (0 binary)
Coefficient statistics:
Matrix range [1e+00, 6e+00]
Objective range [1e+00, 3e+03]
Bounds range [1e+00, 2e+05]
RHS range [1e+00, 1e+00]
Using branch priorities (all 0).
Found heuristic solution: objective 150000.00000
Presolve removed 4365 rows and 276 columns
Presolve time: 0.19s
Presolved: 1255 rows, 2775 columns, 20571 nonzeros
Variable types: 0 continuous, 2775 integer (2719 binary)
Root relaxation: objective 2.500000e+01, 1416 iterations, 0.03 seconds (0.04 work units)
Nodes | Current Node | Objective Bounds | Work
Expl Unexpl | Obj Depth IntInf | Incumbent BestBd Gap | It/Node Time
0 0 25.00000 0 67 150000.000 25.00000 100% - 0s
H 0 0 3036.0000000 25.00000 99.2% - 0s
H 0 0 42.0000000 25.00000 40.5% - 0s
H 0 0 40.0000000 25.00000 37.5% - 0s
H 0 0 33.0000000 25.00000 24.2% - 0s
0 0 25.00000 0 179 33.00000 25.00000 24.2% - 0s
H 0 0 32.0000000 25.00000 21.9% - 0s
H 0 0 31.0000000 25.00000 19.4% - 0s
0 0 25.00000 0 51 31.00000 25.00000 19.4% - 0s
0 0 25.00000 0 100 31.00000 25.00000 19.4% - 0s
H 0 0 30.0000000 25.00000 16.7% - 0s
0 0 25.00000 0 66 30.00000 25.00000 16.7% - 1s
0 0 25.00000 0 135 30.00000 25.00000 16.7% - 1s
H 0 0 28.0000000 25.00000 10.7% - 1s
0 0 25.00000 0 63 28.00000 25.00000 10.7% - 1s
0 0 25.00000 0 63 28.00000 25.00000 10.7% - 1s
0 2 25.00000 0 63 28.00000 25.00000 10.7% - 1s
490 307 25.00000 27 119 28.00000 25.00000 10.7% 148 5s
2060 1090 26.00000 46 70 28.00000 25.00000 10.7% 96.6 10s
3840 1312 25.87500 42 86 28.00000 25.00000 10.7% 97.2 17s
3862 1382 25.87500 43 90 28.00000 25.00000 10.7% 97.6 20s
8784 3562 26.00000 56 85 28.00000 25.00000 10.7% 75.1 27s
11848 5067 26.00000 76 88 28.00000 25.00000 10.7% 75.1 30s
13108 5129 26.00000 42 72 28.00000 25.00000 10.7% 73.4 35s
17547 7511 cutoff 44 28.00000 25.00000 10.7% 73.3 44s
17851 8025 cutoff 38 28.00000 25.00000 10.7% 73.6 45s
19255 8563 26.28376 78 128 28.00000 25.00000 10.7% 74.8 50s
24019 11146 26.66667 61 32 28.00000 25.00000 10.7% 76.1 55s
30054 14593 25.00000 70 103 28.00000 25.00000 10.7% 73.8 68s
32413 14594 26.00000 62 103 28.00000 25.00000 10.7% 73.0 76s
35783 16897 26.00000 50 46 28.00000 25.00000 10.7% 71.7 80s
40976 19210 27.00000 32 27 28.00000 25.00000 10.7% 72.5 88s
42314 20340 26.00000 72 18 28.00000 25.00000 10.7% 72.9 90s
47435 22558 25.22222 93 66 28.00000 25.00000 10.7% 73.4 95s
52079 24566 27.00000 72 67 28.00000 25.00000 10.7% 74.6 100s
57191 27072 27.00000 36 82 28.00000 25.00000 10.7% 75.3 105s
60873 29117 25.00000 33 76 28.00000 25.00000 10.7% 75.8 110s
65007 31139 27.00000 44 65 28.00000 25.00000 10.7% 77.6 115s
69001 32582 25.33333 41 165 28.00000 25.00000 10.7% 78.2 124s
69296 33163 26.00000 42 113 28.00000 25.00000 10.7% 78.3 125s
70625 33387 26.12000 83 73 28.00000 25.00000 10.7% 78.5 131s
73322 35272 cutoff 40 28.00000 25.00000 10.7% 78.9 136s
76051 36623 26.00000 54 79 28.00000 25.00000 10.7% 79.5 140s
79738 38560 25.07242 35 112 28.00000 25.00000 10.7% 80.0 145s
80955 38641 cutoff 42 28.00000 25.00000 10.7% 80.3 153s
81151 39297 cutoff 55 28.00000 25.00000 10.7% 80.3 155s
85169 41709 27.00000 101 35 28.00000 25.00000 10.7% 80.8 160s
89310 43775 26.00000 61 120 28.00000 25.00000 10.7% 80.4 165s
93365 45836 cutoff 75 28.00000 25.00000 10.7% 80.6 170s
96880 47940 26.00000 60 35 28.00000 25.00000 10.7% 80.7 175s
98768 49046 26.00000 45 55 28.00000 25.00000 10.7% 80.4 180s
103461 51606 25.17647 72 75 28.00000 25.00000 10.7% 79.9 185s
104830 51949 cutoff 49 28.00000 25.00000 10.7% 79.8 191s
105452 55120 cutoff 50 28.00000 25.00000 10.7% 79.8 197s
112547 56136 25.02427 56 78 28.00000 25.00000 10.7% 78.7 200s
117465 58233 27.00000 72 102 28.00000 25.00000 10.7% 78.6 206s
118735 58341 26.00000 75 68 28.00000 25.00000 10.7% 78.5 217s
118920 58420 26.50000 98 67 28.00000 25.00000 10.7% 78.6 222s
120481 60031 25.00000 67 117 28.00000 25.00000 10.7% 78.5 225s
125731 61921 cutoff 47 28.00000 25.00000 10.7% 78.5 230s
130745 64169 25.40625 70 117 28.00000 25.00000 10.7% 78.5 236s
135598 66321 26.45455 98 57 28.00000 25.00000 10.7% 78.5 242s
139813 67880 26.75000 71 67 28.00000 25.00000 10.7% 78.3 245s
144289 70071 cutoff 63 28.00000 25.00000 10.7% 78.5 250s
148923 72174 25.50000 64 115 28.00000 25.00000 10.7% 78.5 256s
152857 74075 27.00000 47 109 28.00000 25.00000 10.7% 78.6 260s
157594 75672 26.00000 60 53 28.00000 25.00000 10.7% 78.9 275s
160315 77259 cutoff 62 28.00000 25.00000 10.7% 79.1 280s
165403 79406 27.00000 73 80 28.00000 25.00000 10.7% 79.4 286s
169432 80463 26.00000 43 68 28.00000 25.00000 10.7% 79.5 301s
172144 82236 cutoff 55 28.00000 25.00000 10.7% 79.5 306s
176257 84172 27.00000 92 50 28.00000 25.00000 10.7% 79.5 310s
180049 85912 cutoff 89 28.00000 25.00000 10.7% 79.5 315s
184022 87167 26.50769 79 81 28.00000 25.00000 10.7% 79.5 323s
184284 88016 26.60000 80 78 28.00000 25.00000 10.7% 79.5 325s
188426 89743 26.00000 60 62 28.00000 25.00000 10.7% 79.4 331s
191959 91392 cutoff 52 28.00000 25.00000 10.7% 79.5 336s
194366 92562 cutoff 86 28.00000 25.00000 10.7% 79.6 340s
198255 94642 25.00000 62 101 28.00000 25.00000 10.7% 79.8 345s
203564 97183 26.04688 61 62 28.00000 25.00000 10.7% 79.8 351s
206165 97944 26.00000 55 67 28.00000 25.00000 10.7% 79.7 359s
206415 98385 26.00000 56 82 28.00000 25.00000 10.7% 79.7 360s
209818 99878 cutoff 57 28.00000 25.00000 10.7% 80.0 365s
212351 100745 25.55556 72 63 28.00000 25.00000 10.7% 80.1 373s
213648 102229 27.00000 100 34 28.00000 25.00000 10.7% 80.1 376s
217712 104069 27.00000 69 44 28.00000 25.00000 10.7% 80.0 381s
220860 105299 cutoff 63 28.00000 25.00000 10.7% 80.2 390s
223788 107535 26.00000 55 54 28.00000 25.00000 10.7% 80.1 395s
228404 109842 27.00000 107 75 28.00000 25.00000 10.7% 80.0 400s
232557 111962 27.00000 68 23 28.00000 25.00000 10.7% 80.0 406s
233664 112349 26.00000 80 65 28.00000 25.00000 10.7% 80.0 415s
236927 114156 cutoff 120 28.00000 25.00000 10.7% 80.0 420s
240605 116308 25.53643 50 117 28.00000 25.00000 10.7% 80.2 425s
244884 118263 cutoff 89 28.00000 25.00000 10.7% 80.1 431s
248630 120150 26.83333 49 81 28.00000 25.00000 10.7% 80.0 435s
253612 122524 cutoff 90 28.00000 25.00000 10.7% 79.8 441s
256816 124450 27.00000 82 66 28.00000 25.00000 10.7% 79.9 445s
260735 126476 27.00000 85 48 28.00000 25.00000 10.7% 79.8 450s
262118 126484 26.00000 58 69 28.00000 25.00000 10.7% 79.8 462s
262126 126543 26.00000 59 41 28.00000 25.00000 10.7% 79.8 467s
263772 127967 25.00000 52 60 28.00000 25.00000 10.7% 79.8 471s
267565 129558 25.08387 59 114 28.00000 25.00000 10.7% 79.7 475s
271368 131690 cutoff 86 28.00000 25.00000 10.7% 79.7 480s
Cutting planes:
Gomory: 11
Projected implied bound: 25
MIR: 76
StrongCG: 24
Flow cover: 287
Zero half: 205
Relax-and-lift: 192
Explored 273599 nodes (21843918 simplex iterations) in 482.24 seconds (408.41 work units)
Thread count was 8 (of 8 available processors)
Solution count 9: 28 30 31 ... 150000
Solve interrupted
Best objective 2.800000000000e+01, best bound 2.500000000000e+01, gap 10.7143%0 -
You are setting gurobi_options twice:
option gurobi_options 'timelimit=100 outlev=1 mipgap=0 mipfocus=3';
option gurobi_options 'logfile=C:\AMPL\2023_01_29\Replication01\logfile.temp';The second statement resets gurobi_options to 'logfile=C:\AMPL\2023_01_29\Replication01\logfile.temp', so all of the options set by the first statement are lost. Instead, use one AMPL option statement to specify all of the options for Gurobi. To avoid having a very long line, you can specify more than one option string, and they will be concatenated:
option gurobi_options
'timelimit=100 outlev=1 mipgap=0 mipfocus=3 '
'logfile=C:\AMPL\2023_01_29\Replication01\logfile.temp';Be sure to put a space at the end of the first string (as shown) or at the beginning of the second string.
1 -
You are correct. It is working now. Thank you so much, and I appreciate the quick responses.
0 -
Hello, I am using this code and I need the
_ampl_elapsed_time + _total_solve_elapsed_timeto be between 300 and 600 seconds. How can I do that?# ----------------------------------------
# GILMORE-GOMORY METHOD FOR
# CUTTING STOCK PROBLEM
# ----------------------------------------
reset;
option solver gurobi;
model cut_1_A.mod;
data cut_1_A.dat;
#definizione del problema Cutting_Opt
#NB: rilassamento del vincolo di interezza delle variabili
problem Cutting_Opt: Cut, Number_Cut, Fill_Orders;
option relax_integrality 1;
#definizione del problema Pattern_Gen
problem Pattern_Gen: Use, Reduced_Cost, Width_Limit;
option relax_integrality 0;
# Variabili e parametri per il metodo Gilmore-Gomory
let nPAT := 0;
for {i in FINALTYPES} {
let nPAT := nPAT + 1;
let nbr[i,nPAT] := floor(roll_width/width[i]);
let {i2 in FINALTYPES: i2 <> i} nbr[i2,nPAT] := 0;
};
param tt;
let tt := 0;
repeat {
solve Cutting_Opt;
print Number_Cut;
let tt := tt + _total_solve_elapsed_time;
let {i in FINALTYPES} price[i] := Fill_Orders[i].dual;
solve Pattern_Gen;
let tt := tt + _total_solve_elapsed_time;
if Reduced_Cost > 0.00001 then {
let nPAT := nPAT + 1;
let {i in FINALTYPES} nbr[i,nPAT] := Use[i];
}
else break;
};
# Opzioni per la visualizzazione dei risultati
option display_width 1000;
display nbr;
option display_width 132;
display Cut;
display nPAT;
print Number_Cut;
print "Tempo di esecuzione:";
display _ampl_elapsed_time + _total_solve_elapsed_time;
0 -
You seem to be asking a new question about time limits, which is only indirectly related to the previous posts in this thread. Also, your question is more about how to use AMPL than how to use Gurobi. Thus I recommend submitting your question instead to the AMPL forum at https://discuss.ampl.com.
0
Please sign in to leave a comment.
Comments
7 comments